The Future of Shopping: AI Commerce

The Future of Shopping: AI Commerce | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article The Future of Shopping: AI Commerce Vanderlei Reis, Isabela Godoy This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8958644/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract As Generative Artificial Intelligence (GenAI) like ChatGPT introduces retailing features such as product search, recommendation and checkout, there may be significant impact to retail, e-commerce, digital marketing, and search engines (SE). This study aims to foresight impacts of GenAI applications in such areas, especially e-Commerce. Speculative scenarios derived from signals covering the impact of GenAI on e-commerce, SE, SEO/SEM, shopping assistants, customer purchasing intention, human digital twins, and Large Language Models (LLM) monetization. A 2x2 Scenario Planning Matrix (SPM) speculates how LLM companies (LLMCs) may monetize through advertising or direct sales versus users’ data trust levels. The SPM proposes a probable "AI Commerce” future where GenAI replaces SE and traditional e-commerce for all online purchasing. Other probable futures include GenAI advertising monetization and potential LLM acquisitions by SE companies. A preferable future with enhanced customers data privacy control is proposed. Diegetic Prototypes (DPs) illustrate probable and preferable futures. Speculative scenarios and DPs help retailers and marketplaces understand GenAI's potential to reshape e-commerce, digital marketing, SEs and online consumer behaviour overall. Figures Figure 1 Figure 2 1. Introduction As consumers rely on AI-based search, organic web traffic is reduced (Sommerfeld, 2025), impacting both SE Optimization (SEO) and SE Marketing (SEM). As Artificial intelligence (AI) advances, consumers are encountering new modes of interaction with retailers, such as chatbots using GenAI, which can be understood as Conversational AI (ConAI). OpenAI’s ChatGPT (Peters, 2025 ), Perplexity AI (Capoot and Rooney, 2025 ) and Google’s Gemini (Mehta, 2025 ) are examples of ConAI that introduced features to enable product search, shopping and checkout which may substantially impact retail. As consequence, e-commerce and marketplaces may also suffer a reduction in traffic and conversion from traditional SEs. Purchasing intention when using ConAI is correlated to its ease of use (Jan et al., 2023 ; Lopes et al., 2024 ), usefulness, trust (Elsayad and Mamdouh, 2024 ; Mari et al., 2024 ; Selter et al., 2023 ; Shahzad et al., 2024 ; Silva et al., 2023 ), perceived risk, anthropomorphism, and intelligence (Mpinganjira et al., 2024 ; Sohn et al., 2025 ; Song et al., 2022 ). Enhancements to ConAI may increase purchase intention, leading to more users shopping and buying through it. Concepts such as human digital twins (Barricelli et al., 2019 ; Guo and Lv, 2022 ; Wang et al., 2024 ), digital twin of a customer (Gartner, 2023) and machine customers (Scheibenreif and Raskino, 2023 ) may evolve as new AI models are developed. ConAI would embed such concepts to better assist its users in searching and buying retail products. Be it through advertising products from e-commerce and marketplaces or having merchants to sell directly through ConAI, LLMCs such as OpenAI may explore take rate fees as an additional revenue stream besides subscription models, as currently LLMCs are not profitable (Wiggers, 2025 ) and need to monetize. All these social, technological, and economic signals point to a future where zero-click search for products and frictionless buying through AI shopping assistants becomes the norm for ConAI users. This article attempts to understand what happens to e-commerce and digital marketing as we know with the rise of ConAI based on LLMs. Through Speculative Design (SD) techniques (Dunne and Raby, 2013 ), more specifically SPM (Rhydderch, 2017 ), it is possible to forecast possible futures as well as extrapolate DPs using Design Fiction techniques (Bleecker, 2009 ) that depict such futures where advertising and commerce are greatly impacted. Hence, the following research question was formulated: RQ. Can ConAI replace e-Commerce? As such, this article contributes to online retailing and retail management practice by offering an overview of future possibilities of ConAI in the industry. 2. Methodology According to Dunne and Raby ( 2013 ), SD is a way to explore complex problems and possible futures through imagination and questioning the status quo . This approach to design flourishes through imagination, seeking to shed new light on so-called wicked problems, foster dialogue and debate about different ways of living, and spark and nurture people’s creative thinking (Dunne and Raby, 2013 ). SD serves to materialize the multitude of potential futures our world could inhabit. While it is traditionally understood that the present is determined by historical events, it is equally valid to consider how it may be shaped by projections of the future—by our collective hopes, ambitions, and imagined possibilities (Dunne and Raby, 2013 ). Through SD techniques, we forecast future scenarios with the purpose of using these ideas as tools to better understand the present and to serve as a discussion point for future analysis. To enable the forecast of such future scenarios, we investigate signals from different sources, such as academic papers and news reports. For academic papers, a literature review was done using different search strings within the Scopus database. To understand signals on GenAI and e-Commerce, a search combining “Artificial Intelligence” AND “e-Commerce” was used. To understand purchase intention when using GenAI, a search string combining “purchase intention” AND “artificial intelligence” was used. And to understand the impacts of GenAI personalization, a string with “digital twin” AND “human”. Results were further refined using filters for papers in the timeline of 2019 to 2025 and an exclusion criterion based on relevance to the theme of this research. We also collected news and LLMCs’ announcements on advancements of generative AI solutions into retail, especially e-Commerce and digital marketplace practices, as well as news and reports on the impact of GenAI searching into SEO/SEM. Based on this collection of signals, we extrapolated possible future scenarios using SD techniques. The scenarios are built on a what-if approach (Dunne and Raby, 2013 ) under a SPM (Rhydderch, 2017 ). By gathering and clustering signals, we can plot the x and y axes on the SPM and use that planning technique to come up with what-ifs under the parameters established for each of the four spaces formed by the intersection of both axes. In the possible futures section, we will further explain the context sustaining our speculative scenarios. Under each of the four spaces, it is possible to create several different speculative scenarios. For the purpose of this research, we’ll briefly name all the scenarios that were speculated for the SPM but focus only on the ones that were deemed to be the most probable and preferable, as per the Probable, Plausible, Possible, Preferable - PPPP standard (Dunne and Raby, 2013 ). To materialize those chosen scenarios, we’ll combine the two main speculations (probable and preferable) with a Design Fiction approach: DPs – fictional pieces inserted into that future scenario (Bleecker, 2009 ). 3. Signals Collected This section collects recent news and academic papers representing relevant signals to this research. I. GenAI in e-Commerce Research on AI in retail focuses primarily on recommender systems, especially for e-Commerce. Guha et al . (2021) indicates that AI helps by making omnichannel and mobile shopping more profitable, notably by sharpening personalized recommendations. Advanced AI algorithms enable e-commerce platforms to interpret external data, learn from it, and improve recommendation quality through flexible adaptation (Bawack et al ., 2022). Personalized recommendation systems have become a vital tool for online retailers, enhancing user experience through tailored product suggestions based on individual preferences and behaviours (Dharwadkar and Kannadaguli, 2024). GenAI offers e-commerce capabilities to synthesize a large amount of customer-provided data (prompts), supporting consumers in their shopping journeys with new retailing experiences (Al-Shaikh et al ., 2024). GenAI capabilities span multiple areas: product description and content generation, product image generation, customer support chatbots, personalization and recommendation, product review analysis, and marketing and advertising (Ghaffari, 2024). ConAI in e-commerce naturally answers shopping queries, recommends relevant products, finds product descriptions, and links them to platform products. ChatGPT-driven chatbots provide personalized, interactive experiences improving efficiency and value proposition, offering higher responsiveness and accuracy, reducing consumer effort and time while providing product knowledge (Yue and Basha, 2024). Pantano and Pizzi (2020) showed an increasing number of chatbot patents in the past 20 years, suggesting the use of new conversational agents based on natural language might lead to better interactions between consumers and retailers. The intersection between AI and customer experience in retailing shows chatbots and voicebots represent most researched AI technologies. Studies highlight that chatbots maintain customer interest through convenience, ubiquity, interactivity, technology readiness, ease of use, satisfaction, and trust (Peruchini et al ., 2024). II. GenAI replacing SEs Initial research suggested ChatGPT-like technology would not replace commercial SEs like Google and Bing, expecting LLMs to become SE components serving as additional query options (Strzelecki, 2023). Since OpenAI's LLM integration into Microsoft Bing and Google's AI mode introduction, widespread speculation emerged about ConAI becoming "answer engines" replacing traditional SEs. Conversely, others argue SEs aren't facing obsolescence, expecting search and chat functionalities to coexist harmoniously (Barnard and Artz, 2023). ChatGPT is used by students and industry as a SE to enhance work and research scope, quality, and creativity with daily or weekly frequency (MacLachlan et al. , 2024). Such ConAI use increases despite research showing users trust ChatGPT less than Google and Wikipedia (Jung et al ., 2024). Barnard and Artz (2023) note that the future of search marketing will depend on how search and answer engines understand brand identity, evaluate brand credibility, and deliver and prioritize brand content. The Fall of SEO With ConAI replacing traditional SEs, there are impacts to search marketing and online advertising (Gartner, 2023). SEO, the traditional set of techniques to increase visibility in organic SE results, is also being impacted, with metrics such as click-through rate (CTR) metrics reducing drastically (Sommerfeld , 2025). Recent survey from Bain & Company (Sommerfeld et al. , 2025) finds that about 80% of consumers now rely on GenAI overviews delivered by traditional SEs (for example, Google and Bing) at least 40% of the time, reducing organic web traffic by an estimated 15% to 25%. As a user may perform no clicks after reading an AI overview on Google or after a reply from ChatGPT, this type of search can be called a zero-click search, which impacts both SEO and SEM. A study conducted by Seer Interactive (Haigler, 2024) shows that organic CTR declined about 70% when an AI overview was present on the SE results page. A similar study analysed by Ahrefs (Law and Guan, 2025) shows 34.5% lower average CTR for the top-ranking page, while Amsive (Guevara, 2025) points to a CTR decline of -15.49% on average. Even though these studies show different decline rates for CTR, all of them point to a reduction in CTR. CTR is one of the most important metrics in both SEO and SEM, with the earlier improving ranking of a page on SE results and the later improving ranking amongst sponsored links. SEO services are forecasted to reach US$ 106,9 billion in 2025 (Research and Markets, 2025) while SEM spending is forecasted at US$351.55 billion in 2025 (Statista, 2025). Such market sizes show that the growth of AI assistants based on LLM may have while reducing CTR metrics. Corroborating these studies, Gartner (2023) expects that by 2026 traditional SE volume will drop 25%, as SEM will lose market share to AI chatbots and other virtual agents. Bain (Sommerfeld et al ., 2025) comments on the rise of LLMs for search pointing that roughly 40% to 70% of LLM users use the platforms to conduct research and summarize information (68%), understand the latest news and weather (48%), and ask for shopping recommendations (42%). Adobe Analytics data shows that consumers are already using ConAI to get product recommendations (Pandya, 2025). This data shows that ConAI traffic to United States retail sites has increased by 1.300% between November and December 2024 and the same period in 2023. This trend seems to have persistence, as traffic increased by 1.200% between July 2024 and February 2025. Additional ConAI solutions are expected to come to market from companies such as Apple, Samsung, Amazon and Meta, as well as startups and will be embedded in smartphones and other devices by default, providing further search substitutions, personalization and ease of use (Gartner, 2023). As ConAI users may interact only with the AI itself, with no clicking through additional pages, further adoption of ConAI may impact not only web page traffic, but e-Commerce as well. III. GenAI as shopping assistant Matos et al . (2025) points that as GenAI advances, consumers are encountering new modes of interaction with retailers, such as ConAI. Gupta and Mukherjee (2024) suggest that GenAI search tools can help make the search process simple, convenient, less effortful, and less time consuming, providing better search experiences and improving the chances of arriving at the best purchase decision, showing that GenAI can potentially disrupt consumer’s online search behaviour. Huh and Kim (2024) also indicate that consumers perceive ConAI as and efficient and convenient decision-aid tool for smart shopping, highlighting utilitarian features such as speed and convenience for customers as well as hedonic values as an emotional companion for shopping. Recent examples of customer facing ConAI include Alexa, Amazon’s voice assistant which interacts with customer primarily via voice and is increasingly used for online shopping (Guha et al. , 2021). Lately, as ConAI gains traction with users, LLMCs introduce new features include product search and recommendations as well as shopping and checkout. Perplexity AI introduced e-Commerce features for retail shopping, including the ability to check out right on their ConAI app or website (Perplexity, 2024). These features allow users to take a photo of an item to search for products as well as to see traditional product cards when users search through shopping questions. Product recommendation is not sponsored, with expectation of unbiased search results. Open AI has also gradually introduced features to enable e-Commerce, previewing new ChatGPT features for users to do research around product showing product cards with images, prices and ratings (Peters, 2025). As hyperscalers such as Google and Amazon are developing their own LLMs and ConAI applications, e-Commerce-like capabilities are also being explored by those. Google has announced agentic checkout and shopping features to be embedded on its SE. New shopping features are also being introduced to Google’s AI Mode, allowing users to engage in a conversation with considerations on product reviews, prices, colour options and availability. Google’s AI Mode will enable virtual try-ons for clothes and agentic checkout to track prices for users (Rincon, 2025). Amazon has introduced ConAI-features to its platform through an expert shopping assistant called Rufus (Mehta and Chilimbi, 2024) to answers users’ questions on shopping needs, products and comparisons and make recommendations. Amazon has also launched a feature called “Buy for Me” (Amazon, 2025), leveraged through agentic AI capabilities, allowing customers to purchase products directly from other websites IV. Purchase Intention through ConAI Jan et al . (2023) found four major reasons for using ConAI for retailing. These are perceived usefulness, perceived ease of use, perceived trendiness, and perceived informativeness, which have significant positive effects on attitudes toward the usage of ConAI. Jan et al . (2023) also discusses three major reasons against ConAI for shopping. These are the usage barrier, functional risk barrier, and intrusiveness barrier for ConAI. Song et al. (2022) found that perceived communication quality and privacy risk mediate the effect of ConAI on adoption intention. While AI personalized services increase convenience while increasing users’ privacy concerns, Song et al. (2022) also perceives that consumers’ willingness to accept ConAI is enhanced if they perceive lower privacy risks during the interaction. However, Sohn et al. (2025) find that consumers express lesser privacy concerns in the presence of AI (vs. human) service agents. This effect stems from consumers’ perceptions that AI agents have less power than human agents. ElSayad and Mamdouh (2024) identifies trust as critical in customer interactions with smart technologies, particularly in e-commerce where face-to-face interaction is absent. Customers rely on trust in platform reliability, security and integrity, shaping perceptions and decisions throughout online shopping. Trust fundamentally impacts perceived service value and purchase decisions. Customer technology readiness influences trust in AI-based retail services through motivators (optimism and perceived innovativeness) and inhibitors (discomfort and insecurity). Technology readiness significantly predicts perceived trust, subsequently influencing perceived usefulness and purchase intentions. Silva et al. ( 2023) also indicate that user trust significantly predicts behavioural intention toward ConAI, while also reducing perceived risk and enhancing flow experience. Huh and Kim (2025) have found that utilitarian and technological value are more important than hedonic and social value. As such speed and convenience are central to retailing through ConAI. On the hand, Selter et al . (2023) found no significant results for the utilitarian and hedonic product types, but rather a decreasing usage intention for automated shopping processes. In addition, Selter et al . (2023) found that trust and behavioural control mediate the effect of automation on usage intention, and this effect is further moderated by inherent novelty seeking. Shahzad et al. (2024) found that trust is tied to customers’ perceptions of the privacy and security of their data during interactions with the ConAI. Lopes et al . (2024) shows AI in online retailing significantly alters consumer buying behaviour through personalized recommendations, simplifying decision-making and increasing purchase likelihood. Digital shopping behaviour is increasingly influenced by psychosocial factors shaping purchase intention: subjective norms (community expectations influencing AI attitudes), faith (trust in AI), consciousness (ethical awareness) and perceived control (belief in controlling AI). When these factors favourably impact AI-enabled ease of use, purchase intention is perceived favourably. Mpinganjira et al . (2023 found that the continued use of chatbots is primarily driven by experiential utilitarian value, followed by hedonic value. However, perceived ConAI risk negatively influences customer experience value in retailing due to concerns about exposing sensitive personal information to unauthorized users. Mari et al . (2024) determines correlation between trust, usefulness and perceived empathy in ConAI and its use as decision aid assistant, delegated agent or trusted product recommender. Opposing such studies, Aiolfi (2023) found that perceived privacy risk, innovativeness and social attraction have been found to not significantly impact attitudes towards ConAI, in particular smart speakers. One possible justification might come from the Privacy Paradox: although people say they care about their privacy and are not willing to share their data, actually they give their private information in exchange for small benefits or convenience. V. Digital Twin of a Customer Digital twins are computer-based models simulating physical entities—objects, processes, humans or human features (Barricelli et al ., 2019). Wang (2024) explores human digital twins that reflect multi-dimensional human information—physical, psychological, and cognitive data—enabling modelling and simulation of mental workload, interactions and trust to replicate human decision-making roles. In retail, Guo and Lv (2022) examines digital twins for user-product interaction in offline and online sales, enabling shoppers to experience goods through virtual reality using digital twins of users and products. Gartner (2023) suggests that digital twins of a customer will allow for simulation and anticipation of customer behaviour. Yang et al . (2025) shows that digital twins have been used in retail to provide virtual replicas of products, service systems, shopping environments and customers interactions allowing consumers to experience products in real-time and reduce product misfit and return losses. Scheibenreif and Raskino (2023) predicted machine customers would outperform humans at information searching, deal negotiation, purchasing, and other buying processes, only requiring trust in both technology and privacy for larger consumer adoption. VI. LLMCs Monetization Problem LLMCs have raised significant resources and have large valuation figures, with Open AI funding at US$ 40 billion, Perplexity AI at US$ 14 billion and Anthropic at US$ 3.5 billion. Despite such figures and growing monthly users, it seems these companies are yet to become profitable (Wiggers, 2025). For instance, ChatGPT’s Open AI reported US$ 5 billion loss in 2024, despite US$ 5.5 billion in revenue from subscriptions (Field, 2025). In a not-so-distant past, big tech companies such as Google and Facebook started monetizing by capturing advertising revenue (Curran et al ., 2011). Google generates revenue by monetizing its SE services through ads services, where advertisers pay money to have their listing ranked higher in search results (KRITZINGER, 2013). Most SEs offer pay-per-click practices to generate revenue (Seymour et al ., 2011). Facebook offers a model to advertise and engage with consumers (Curran et al. , 2011). Instagram also relies on advertising to monetize (Burgess et al. , 2024). It is not farfetched to imagine that LLMCs will try and capture advertising revenue, besides monetizing through a subscription model. 4. Possible Futures: 2x2 SPM Collected signals perceived in social, economic, and technological spaces were used to support defining axes in the 2x2 SPM. It is clear that LLMCs need to increase their revenue and may choose to do so through a stream different from the existing subscription model. LLMCs are investing in capabilities for product recommendation and one-click checkout, which may lead to introduction ads in their ConAI. LLMCs may choose to sell merchants’ products, similarly to marketplaces. These choices are reflected in the SPM. We define the vertical axis of SPM by poising the question “How will LLMCs monetize besides subscriptions?” as seen in Figure I. As variables in this axis, answering the questions, we propose the choice of “Selling” and “Advertising”. Evaluating customer purchase intention through ConAI, easy usage and trust are key factors for users to shop through an AI. As we consider easy usage a matter of time and investment, trust is the real psychological factor that may decide whether ConAI will succeed in either showing product ads or selling merchant products. As we understand trust may be a concept difficult to grasp as a product feature that a LLM company can embed on a ConAI, we consider user data a better simulacrum for trust. So, we chose to poise the question “Do people trust their own data to AI assistants?” On the horizontal axis of the SPM, as seen in Figure I. Add Figure I here We speculate four speculative scenarios, as seen on Figure I: Speculative Scenario 1: LLMCs monetize through selling and consumers trust their data uncritically to ConAI LLMCs invest in shopping and checkout capabilities embedded in their ConAI sites and apps. As consumers continue to provide personal data to ConAI, these solutions end up becoming digital personal shoppers through agentic AI features and evolve into human digital twins as LLMs become capable of representing virtually any consumer. LLMCs introduce API communication to facilitate any vendor or merchant to add product details such as description, specs, photos, videos, price and stock availability so ConAI can better recommend products to their users. LLMCs charge a take rate fee for every merchant sale to consumers through ConAI sites or apps. As buying through ConAI becomes the standard, traditional retail e-commerce and marketplace platforms end-up as obsolete. In this scenario, we speculate that AI commerce, where consumers purchase goods and services through ConAI, becomes the standard for any online purchasing. Speculative Scenario 2: LLMCs monetize through selling, but consumers are sceptical about trusting their data to ConAI LLMCs monetize selling goods and services for other vendors and merchants through ConAI sites and apps. In this scenario consumers are sceptical to trust their personal data to a ConAI, not willingly providing demographic, behavioural, and transactional data. With this limitation, product search and recommendation are not user-oriented, but rather generic, with no personalization. As perceived ease of use and trust are suboptimal in this scenario, purchase intention is low (Lopes et al ., 2024). Therefore, sales figures through ConAI are also low. We speculate that in this scenario, only geeks and early adopters end up using ConAI for shopping and checkout. We speculate that LLMCs will implement other strategies to increase perceived trust, such as brand partnerships to increase sales by offering discounts for group sales in bulk, as well as loyalty programs for their users, offering cashback. However, we speculate that such tactics will fail and LLMCs will end up not breaking even, which would force them to sell to hyperscalers to avoid bankruptcy in the long run. Speculative Scenario 3: LLMCs monetize through advertising and consumers trust their data uncritically to ConAI LLMCs chose to increase their revenue by pursuing advertising budgets from other companies while consumers feed ConAI with personal data uncritically. With such data, ConAI solutions can offer hyperpersonalized product recommendations, which deliver greater CTR and conversion to e-commerce and marketplace sites that advertise through the ConAI. This evolution renders SEO and SEM completely obsolete. In addition to ads revenue, LLMCs continue to collect revenue from ConAI subscriptions. Following trends observed in other subscription models, LLMCs introduce tiers to ConAI subscription, including an ads free premium tier. Speculative Scenario 4: LLMCs monetize through advertising, but consumers are sceptical about trusting their data to ConAI LLMCs chose advertising as new business model, however consumers do not trust their personal data to ConAI. This leads ConAI to provide unbiased product recommendations to its users. We speculate that this will make ConAI accepted but not used to a degree where SEs may completely disappear. As such, SEO and SEM continue to evolve, further developing features such as AI Overview and AI Mode. With continued competition with traditional SEs, now fully embracing AI, LLMCs do not break even and end up acquired by hyperscalers, as LLMs replace traditional search indexers, becoming more of a relevant tool than a complete solution. 5. Diegetic Prototypes a. Most Probable Future (Speculative Scenario 1) As the most probable future, we pictured a world where AI Commerce is present in everyone’s life one way or the other. In that world, a Podcast called “Checkout” becomes popular. Here is a transcript of one the show’s episodes. “Checkout Podcast Season 3 Episode 11 - The Gallery” Opening: You’re listening to Checkout — the number 1 retail-tech podcast for buyers and sellers. Bela: Today, we are checking out the stories behind a new luxury market that quietly emerged while we were all still online shopping. And joining me is our beloved Vander. Vander : Thanks, Bela. Out of Home is no longer just a media strategy term; it’s become the latest hype among high-ticket consumers. But before we dive deeper, we’ve got a special guest with us today, right? Bela : That’s right! Our dearest digital influencer, Miquela, a.k.a. “@lilmiquela”, is here with a message from OpenBrain. Miquela : Hi Checkoutters! OpenBrain has an amazing update – I can be your digital twin for FREE! Subscribers, just login and select "MyQuela." Customize your shopping preferences and create a MyQuela that looks like you. I automate recurring purchases, find the best deals, and send you releases matching your style. Browse OpenBrain while you listen – I handle all payments. Non-subscriber, stick around for a gift! Bela: Thanks, Miquela. I already have my digital twin with MyQuela, and what a joy! Everything arrives without me even having to think about it. So practical, so fast. Vander: Not to mention the savings! I’ve been a subscriber since MyQuela launched the twin feature back in 2030, and I noticed the difference in my budget right away! It finds exactly what I want, at the best price and highest quality. Bela: Exactly! And look, while you and I are using our digital personas to speed up shopping, there’s a crowd who wants to live the full in-person shopping experience. And that’s who we’re talking about today. Vander : That’s right. Today, we’re talking about the new offline trend in the luxury market. The Gallery is a new Shopping World with stunning landscapes, top-notch hospitality, diverse brands, and eclectic cuisine. Bela : Shopping World? Don’t you mean shopping malls? Vander : laughs. No, Bela. I really do mean World. Here’s the deal: since 2025, even before digital twins and AI shopping tools became mainstream, luxury brands had already noticed two big trends: a strong focus on longevity and the need to sell experiences more than just products. That’s how we started seeing initiatives like Prada Caffè, Bar Luce, Gucci Osteria, The Tiffany Blue Box Café, the Louis Vuitton restaurant... Bela: I remember when all that started! Now we have that luxury Ralph Lauren hospital in São Paulo and that Tommy Hilfiger Speed Track! I’ve always wanted to try it. Vander: Oh, it is amazing. I celebrated my birthday there last month, and it was awesome. And get this: Yang Huiyan picked up on this trend and launched the first Shopping World in Shenzhen in 2036. Imagine a theme park, but for adults who love luxury brands and have the time to do in-person shopping. Bela : Yang Huiyan, you mean the Real Estate Chinese Billionaire? Vander : Precisely. She had the means and was deeply involved with her audience to start this new business. Bela: I mean, it does make sense. And I am assuming that the name for her new enterprise is no coincidence. The Gallery. I bet it has something to do with her background in the arts. Vander : You are absolutely right. She is a Bachelor of Arts/Science by Ohio State University, and that influenced her decision for the name and the whole business model. Her Western experience gave her a broader understanding of appealing to both Westerners and Easterners. The artistic philosophy behind The Gallery’s success: every product has to be fine-art-like, and every purchase should be special. Bela : So, you are not just taking a bag home, you are purchasing a feeling. Brilliant! But, can you imagine the ticket prices?! *laughs* Vander : That was another genius move: no tickets — invitation only. Total exclusivity. For The Gallery’s first edition, she partnered with 45 brands that had to design unique experiences specifically for that World. Patek Philippe, for example, built a giant watchmaking workshop where visitors could create their own watch straps. A brand that values tradition and longevity created a space where parents and kids could share a memory that would last forever and keep perfect time too. Bela : That’s incredible! It makes sense she would target only the richest. Who else got the time to do in-person shopping in this economy? Vander : Right? The current scenario is: we’ve still got folks shopping on the web using their favourite LLM (the late AI adopters). Then the growing crowd using human digital twins. And finally, there’s this luxury crowd using their time and money traveling abroad in search of offline shopping experiences. Bela : Funny how you didn’t even mention the old SEs *laughs*. I don’t think I’ve opened one for shopping since 2030. Do any still exist? Vander: There are still a few survivors on life support. Almost no one uses a browser for searching — let alone shopping — especially with so many free conversational models available. Bela : And I also haven’t seen any in-person shopping/market/mall. There are some small garage sales, but since no one has the time to go places and do in-person purchases, we’ve all been online shopping with drone-dropped deliveries. Vander: Right? It’s not about money anymore. No one has the time to drive far away just to pick some apples and oranges. Let AmazonFresh handle that. Bela : Amazon, if you are listening, we’d love to partner with you *laughs*. I still think it’s crazy that those giant tech companies from the past either died or had to completely reinvent themselves. Vander: Totally. Last year, 2039, I saw the last major bankruptcy. Now, the old SEO teams are trying to grab a slice of LLM product recommendation space and cramming in Ads wherever they can. But for premium subscribers, it’s all fully personalized, zero ads. It’s amazing. Bela: Smart were the tech companies that pivoted early. Besides the premium subscriptions; they also earn a fee on every purchase completed through the LLM. Is that true? Vander: It is. For users without purchase profiles, tech companies make money through ads. For those who use digital twins or premium shopping features like automation, they win a fee for each purchase. Bela: Just thinking about all this makes me want to visit that Shopping World. Maybe play soccer with Emirates-sponsored players? *laughs* Vander: While we’re here automating daily shopping and leaving the mall runs to our digital twins, some folks have time to spend days immersed in those incredible experiences. Bela: Who knows, maybe by 2050 I’ll be one of them? Let’s go to Shopping World Brazil edition? Vander: Let’s go! Maybe we’ll land a partnership with Yang Huiyan and score an invite. Until then, digital twins it is. Bela: For sure! As promised, for those of you who stayed till the end, go to OpenBrain use our promo code for the premium plan: CHECKOUT30 . Thanks, Miquela, OpenBrain! Vander: Thanks, everyone. We are checking out. (DP - Developed by the authors) b. Preferable Future As a preferable scenario, we pictured a world where some users approach data trust more critically, with a better balance between ad usage for AI Commerce monetization. On that future, people can choose how much they want to invest in their own privacy, however more control means a more costly option for AI commercial usage, while less control means lower prices for AI usage. This scenario is closer to the intersection point of the two axes in the SPM presented in Figure I. Add Figure II here Under the preferable scenario, as seen in the UI on Figure II, data use is limited. For instance, there is no Digital Twin available for Free plan, with data largely taken from users to enhance ad segmentation, making that plan still profitable for AI companies. Data usage and storage is still limited according to applicable laws and regulations, which users cannot consent to under the free plan and thus makes it impossible for that option to have a digital twin as this would require personal data that cannot be given without consent. The second plan and third plan differ precisely on how much data can be taken from users and how much they can control its use for all AI actions, though these plans are more costly. 6. Conclusion In this article we collected signals that were used to plot a SPM and create DP using SD and Design Fiction techniques. Based on the SPM, we speculated possible futures where LLMCs monetize through advertising and/or selling merchant products directly on ConAI. We speculate whether users would trust their personal data to ConAI uncritically or sceptically. We further explore possible futures, understanding that the most probable future has AI commerce dominating, with ConAI replacing SEs as well as traditional retail e-commerce and marketplaces. We also speculate on a preferable future, where users may choose different subscription tiers on ConAI solutions to better control their personal data as well as exposition to ads. Though this speculative exercise aims to be comprehensive when it comes to ConAI, SEs, LLMCs, human digital twins and psychological factors influencing consumers, it has been limited towards social commerce trends and implications for the labour market and other social impacts. Also, this article does not explore and the future of shopping on physical stores or the impact of AI on physical stores, although DP speculated in this paper reaffirm the domination of AI commerce in retailing and hypothesize a future where all physical retail stores become a luxury. Further exercises may encompass discussion on these areas. Declarations Author Contribution I.G. contributed with the methodology on speculative design while V.R. gathered the signals from academic and non-academic sources. Both authors worked on the building one the possible and probable futures. V.R. created the 2x2 matriz. I.G. created the design fiction pieces. Both authors reviewed the manuscript. Acknowledgement The authors acknowledge methodology guidance on speculative design techniques from Prof. Diogo Cortiz, PhD, theoretical support on artificial intelligence from Prof. Dora Kauffman, PhD as well as support from Bruno Johnson, MS. References Aiolfi S. (2023), "How shopping habits change with artificial intelligence: smartspeakers' usage intention". International Journal of Retail & Distribution Management , Vol. 51 No. 9-10 pp. 1288–1312, doi: doi.org/10.1108/IJRDM-11-2022-0441 Al-Shaikh, M.S., Al-Mousa, M.R., Al-Ababneh, H.A., Ali, M.W., Alkaawneh, S.M., Barhoush, F.M.S. and Binsaddig, R. (2024), “The challenges of using generative AI applications in electronic commerce”, in IEEE (Ed.), Proceedings of the 25th International Arab Conference on Information Technology (ACIT 2024), IEEE, Piscataway, NJ, pp.1–5, doi: doi.org/10.1109/acit62805.2024.10876935 Amazon (2025), “Amazon Shopping App: ‘Buy for Me’ feature enables purchases from other brands’ sites”, available at: https://www.aboutamazon.com/news/retail/amazon-shopping-app-buy-for-me-brands (accessed 25 June 2025). Barnard, J. and Artz, M. (2023), “Search marketing in the age of AI: understanding the marketing implications of search, assistive and answer engines”, Journal of Digital & Social Media Marketing , Vol. 11 No. 3, p.244, doi: doi.org/10.69554/zvmd6442 Barricelli, B., Casiraghi, E. and Fogli, D. (2019), “A survey on digital twin: definitions, characteristics, applications, and design implications”, IEEE Access , Vol. 7, pp.167653–167671, doi: doi.org/10.1109/ACCESS.2019.2953499 Bawack, R.E., Wamba, S.F., Carillo, K.D.A. and Akter, S. (2022), “Artificial intelligence in e-commerce: a bibliometric study and literature review”, Electronic Markets , Vol. 32, pp.297–338, doi: doi.org/10.1007/s12525-022-00537-z Bleecker, J. (2009), Design Fiction: a short essay on design, science, fact and fiction , Near Future Laboratory, Los Angeles, CA. Burgess, J., Carah, N., Angus, D., Obeid, A. and Andrejevic, M. (2024), “Why am I seeing this ad? The affordances and limits of automated user-level explanation in Meta’s advertising system”, New Media & Society , Vol. 26 No. 9, pp.5130–5149, doi: doi.org/10.1177/14614448241251796 Capoot, A. and Rooney, K. (2025), “OpenAI hits $10 billion in annualized revenue fueled by ChatGPT growth”, available at: https://www.cnbc.com/2025/06/09/openai-hits-10-billion-in-annualized-revenue-fueled-by-chatgpt-growth.html (accessed 25 June 2025). Curran, K., Graham, S. and Temple, C. (2011), “Advertising on Facebook”, International Journal of E-Business Development , Vol. 1. Dharwadkar, V., R, V., S, M., M G, J. and Kannadaguli, P. (2024), “Smart cart: revolutionizing e-commerce in India with AI-powered personalized product recommendations overview”, in IEEE (Ed.), Proceedings of the 5th International Conference on Circuits, Control, Communication and Computing (I4C 2024), IEEE, Piscataway, NJ, pp.87–92, doi: doi.org/10.1109/i4c62240.2024.10748521 Dunne, A. and Raby, F. (2013), Speculative Everything: design, fiction, and social dreaming , MIT Press, Cambridge, MA. Elsayad, G. and Mamdouh, H. (2024), “Are young adult consumers ready to be intelligent shoppers? The importance of perceived trust and the usefulness of AI-powered retail platforms in shaping purchase intention”, Young Consumers , Vol. 25 No. 6, pp.969–989, doi: doi.org/10.1108/YC-02-2024-1991 Field, H. (2025), “Perplexity AI wrapping talks to raise $500 million at $14 billion valuation”, available at: https://www.cnbc.com/2025/05/12/perplexity-funding-round-comet.html (accessed 25 June 2025). Gartner (2023), “Predicts 2024: how GenAI will reshape tech marketing”, available at: https://www.gartner.com/en/doc/800771-predicts-2024-how-genai-will-reshape-tech-marketing (accessed 25 June 2025). Ghaffari, G., Yousefimehr, B. and Ghatee, M. (2024), “Generative-AI in e-commerce: use-cases and implementations”, in IEEE (Ed.), 2024 20th CSI International Symposium on Artificial Intelligence and Signal Processing (AISP 2024), Babol, Iran, 21–22 February 2024 , IEEE, Piscataway, NJ, pp.1–5, doi: doi.org/10.1109/AISP61396.2024.10475266 Guevara, W. (2025), “Google AI Overviews: new CTR study reveals how to navigate negative SERP impact”, available at: https://www.amsive.com/insights/seo/google-ai-overviews-new-research-reveals-how-to-navigate-click-drop-off/ (accessed 25 June 2025). Guha, A., Grewal, D., Kopalle, P.K., Haenlein, M., Schneider, M.J., Jung, H., Moustafa, R., Hegde, D.R. and Hawkins, G. (2021), “How artificial intelligence will affect the future of retailing”, Journal of Retailing , Vol. 97 No. 1, pp.28–41, doi: doi.org/10.1016/j.jretai.2021.01.005 Gupta, A.S. and Mukherjee, J. (2025), “Framework for adoption of generative AI for information search of retail products and services”, International Journal of Retail & Distribution Management , Vol. 53 No. 2, pp.165–181, doi: doi.org/10.1108/IJRDM-05-2024-0203 Guo, J. and Lv, Z. (2022), “Application of digital twins in multiple fields”, Multimedia Tools and Applications , Vol. 81, pp.26941–26967, doi: doi.org/10.1007/s11042-022-12536-5 Haigler, N. (2024), “How AI overviews are impacting CTR: 5 initial takeaways”, available at: https://www.seerinteractive.com/insights/how-ai-overviews-are-impacting-ctr-5-initial-takeaways (accessed 25 June 2025). Huh, J. and Kim, H. (2025), “‘Function or emotion?’: understanding smart shopping via voice AI based on the LVSEM approach”, International Journal of Retail & Distribution Management , Vol. 53 No. 7, pp.623–638, doi: doi.org/10.1108/IJRDM-07-2024-0370 Jan, I. U., Ji, S. And Kim, C. (2023) "What (de) motivates customers to use AI-powered conversational agents for shopping? The extended behavioral reasoning perspective", Journal of Retailing and Consumer Services, Vol. 75, doi: doi.org/10.1016/j.jretconser.2023.103440 Jung, Y., Chen, C., Jang, E. and Sundar, S.S. (2024), “Do we trust ChatGPT as much as Google Search and Wikipedia?”, in Mueller, F.F., Kyburz, P., Williamson, J.R. and Sas, C. (Ed.s), CHI 2024 — Extended Abstracts of the 2024 CHI Conference on Human Factors in Computing Systems, Association for Computing Machinery, New York, NY, pp.1–9, doi: doi.org/10.1145/3613905.3650862 Law, R. and Guan, X. (2025), “AI overviews reduce clicks by 34.5%”, available at: https://ahrefs.com/blog/ai-overviews-reduce-clicks/ (accessed 25 June 2025). Lopes, J.M., Silva, L.F. and Massano-Cardoso, I. (2024), “AI meets the shopper: psychosocial factors in ease of use and their effect on e-commerce purchase intention”, Behavioral Sciences , Vol. 14 No. 7, p.616, doi: doi.org/10.3390/bs14070616 MacLachlan, R.J.R., Adams, R., Lauro, V., Murray, M., Magueijo, V., Flockhart, G. and Hasty, W. (2024), “Chat-GPT: a clever SE or a creative design assistant for students and industry?”, in Grierson, H., Bohemia, E. and Buck, L. (Ed.s), DS 131: Proceedings of the International Conference on Engineering & Product Design Education (E&PDE 2024): Rise of the Machines: Design Education in the Generative AI Era, Aston University, Birmingham, UK, pp.414–419, doi: doi.org/10.35199/epde.2024.70 Mari, A., Mandelli, A. and Algesheimer, R. (2024), “Empathic voice assistants: enhancing consumer responses in voice commerce”, Journal of Business Research , Vol. 175, p.114566, doi: doi.org/10.1016/j.jbusres.2024.114566 Mehta, I. (2025), “Google introduces Shop with AI Mode with price tracking, agentic checkout, and virtual try-on”, available at: https://techcrunch.com/2025/05/20/google-adds-ai-powered-shopping-features-for-discovery-and-easy-check-out/ (accessed 25 June 2025). Mehta, R. and Chilimbi, T. (2024), “Amazon announces Rufus, a new generative AI-powered conversational shopping assistant”, available at: https://www.aboutamazon.com/news/retail/amazon-rufus (accessed 25 June 2025). Mpinganjira, M., Dlodlo, N. and Idemudia, E.C. (2024), “Perceived experiential value and continued use intention of e-retail chatbots”, International Journal of Retail & Distribution Management , Vol. 52 No. 13, pp.121–135, doi: doi.org/10.1108/IJRDM-04-2023-0237 OpenAI (2025), “Help ChatGPT discover your products”, available at: https://openai.com/chatgpt/search-product-discovery/ (accessed 29 August 2025). Pandya, V. (2025), “Adobe Analytics: traffic to U.S. retail websites from generative AI sources jumps 1200 percent”, available at: https://blog.adobe.com/en/publish/2025/03/17/adobe-analytics-traffic-to-us-retail-websites-from-generative-ai-sources-jumps-1200-percent (accessed 25 June 2025). Pantano, E. and Pizzi, G. (2020), “Forecasting artificial intelligence on online customer assistance: evidence from chatbot patents analysis”, Journal of Retailing and Consumer Services , Vol. 55, doi:doi.org/10.1016/j.jretconser.2020.102096 Perplexity (2024), “Shop like a pro: Perplexity’s new AI-powered shopping assistant”, available at: https://www.perplexity.ai/hub/blog/shop-like-a-pro (accessed 25 June 2025). Peruchini, M., Da Silva, G.M. and Teixeira, J.M. (2024), “Between artificial intelligence and customer experience: a literature review on the intersection”, Discover Artificial Intelligence , Vol. 4, p.4, doi: doi.org/10.1007/s44163-024-00105-8 Peters, J. (2025), “ChatGPT is getting better for shopping”, available at: https://www.theverge.com/news/656729/openai-chatgpt-search-shopping (accessed 25 June 2025). Research and Markets (2025), “SE Optimization Services Market Report 2025”, available at: https://www.researchandmarkets.com/report/seo (accessed 18 June 2025). Rhydderch, A. (2017), “Scenario building: the 2x2 matrix technique”, in Bourse, F. and Roëls, C. (Ed.s), Prospective and Strategic Foresight Toolbox , Futuribles International, France. Rincon, L. (2025), “Shop with AI Mode, use AI to buy and try clothes on yourself virtually”, available at: https://blog.google/products/shopping/google-shopping-ai-mode-virtual-try-on-update/ (accessed 25 June 2025). Scheibenreif, D. and Raskino, M. (2023), When Machines Become Customers , Gartner Inc., Stamford, CT. Selter, J., Fota, A., Wagner, K. and Schramm-Klein, H., 2023, “Aspects driving customers' intention to use automated purchasing processes” International Journal of Retail & Distribution Management, Vol. 51, No. 9-10, pp. 1158 -1173, doi: doi.org/10.1108/IJRDM-10-2022-0397 Seymour, T., Frantsvog, D. and Kumar, S. (2011), “History of SEs”, International Journal of Management & Information Systems (IJMIS) , Vol. 15 No. 4, pp.47–58, doi: doi.org/10.19030/ijmis.v15i4.5799 Shahzad, M. F., Xu, S., An, X. And Javed, I. (2024), "Assessing the impact of AI-chatbot service quality on user e-brand loyalty through chatbot user trust, experience and electronic word of mouth", Journal of Retailing and Consumer Services, Vol. 79, doi: doi.org/10.1016/j.jretconser.2024.103867 Silva S.C., De Cicco R., Vlačić B., and Elmashhara M.G. (2023), "Using chatbots in e-retailing – how to mitigate perceived risk and enhance the flow experience". International Journal of Retail & Distribution Management, Vol. 51 No. 3 pp. 285–305, doi: doi.org/10.1108/IJRDM-05-2022-0163 Sohn, S., Labrecque, L., Siemon, D. and Morana, S. (2025) “Artificial intelligence versus human service agents: How their presence shapes consumer information privacy concerns”, Journal of Retailing , Vol. 101, No. 2, pp. 263-278, doi: doi.org/10.1016/j.jretai.2025.03.003 Sommerfeld, N., McCurry, M. and Harrington, D. (2025), “Goodbye clicks, hello AI: zero-click search redefines marketing”, available at: https://www.bain.com/insights/goodbye-clicks-hello-ai-zero-click-search-redefines-marketing/ (accessed 25 June 2025). Song, M., Xing, X., Duan, Y., Cohen, J., and Mou, J. (2022) "Will artificial intelligence replace human customer service? The impact of communication quality and privacy risks on adoption intention", Journal of Retailing and Consumer Services , Vol 66, doi: doi.org/10.1016/j.jretconser.2021.102900 Statista (2025), “Search advertising – worldwide”, available at: https://www.statista.com/outlook/amo/advertising/search-advertising/worldwide (accessed 25 June 2025). Strzelecki, A. (2024), “Is ChatGPT-like technology going to replace commercial SEs?”, Library Hi Tech News , Vol. 41 No. 6, pp.18–21, doi: doi.org/10.1108/lhtn-02-2024-0026 Wang, B., Zhou, H., Li, Z., Yang, G., Zheng, P., Song, C., Yuan, Y., Wuest, T., Yang, H. and Wang, L. (2024), “Human digital twin in the context of Industry 5.0”, Robotics and Computer-Integrated Manufacturing , Vol. 85, p.102626, doi: doi.org/10.1016/j.rcim.2023.102626 Wiggers, K. (2025), “OpenAI is losing money on its pricey ChatGPT Pro plan, CEO Sam Altman says”, available at: https://techcrunch.com/2025/01/05/openai-is-losing-money-on-its-pricey-chatgpt-pro-plan-ceo-sam-altman-says/ (accessed 25 June 2025). Yang, Q., Sun, L., Wu, X. and Li, Y. (2025), “Implications of retailer-owned digital twins services: the trade-offs between customer experience, misfit returns reduction, and investment costs”, Journal of Retailing and Consumer Services , Vol. 88, doi: doi.org/10.1016/j.jretconser.2025.104478 Yue, Y., Ng, S. and Basha, N.K. (2024), “Consumption values, attitudes and continuance intention to adopt ChatGPT-driven e-commerce AI chatbot (LazzieChat)”, Pakistan Journal of Commerce and Social Sciences , Vol. 18 No. 2, pp.249–284. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8958644","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":596375823,"identity":"17ce5bc4-a373-4c86-bdda-2fab3da757c3","order_by":0,"name":"Vanderlei Reis","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+ElEQVRIiWNgGAWjYJACCQYGCwY2ECuBwQYixNhAUIsETEsaCVqg4DBhLQbHzx688XGPhBwfA+/DBw/+nM8zOH782QPGHfdwazmTl2w545mEMRsDu7FBYtvtYoMzOeYGjGeKcWqRbMgxk+Y5IJHYxsDGJpHYcDtxww0eNgnGtgTcWvrfmEn/OSBRD9TC/iPhzzmgFvZneLXwSwBtYTggkcAGtIUhge0AUAuDGQEtb4wtew5IGLYxszEDnZdcLAnyS+IZ3FrY+HMMb/w4YCMv397G+PHHH7s8PlCIfdyBWwsCMEMokFI2BmI0wAFEyygYBaNgFIwCJAAAwMNPHJDzexwAAAAASUVORK5CYII=","orcid":"","institution":"","correspondingAuthor":true,"prefix":"","firstName":"Vanderlei","middleName":"","lastName":"Reis","suffix":""},{"id":596375824,"identity":"0964e824-643c-4dba-b9b6-118d2e928f7e","order_by":1,"name":"Isabela Godoy","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Isabela","middleName":"","lastName":"Godoy","suffix":""}],"badges":[],"createdAt":"2026-02-24 14:38:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8958644/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8958644/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":103483194,"identity":"f0c718e6-c766-4c00-9cd0-384c4c8f3edb","added_by":"auto","created_at":"2026-02-26 08:27:27","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":204555,"visible":true,"origin":"","legend":"\u003cp\u003e\u0026nbsp;Legend not included with this version.\u003c/p\u003e","description":"","filename":"FigureI.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8958644/v1/c4c598b991374999085e6289.jpg"},{"id":103483271,"identity":"c39eddf0-4c0d-4de5-acb7-cb607c057461","added_by":"auto","created_at":"2026-02-26 08:27:38","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":113830,"visible":true,"origin":"","legend":"\u003cp\u003e\u0026nbsp;Legend not included with this version.\u003c/p\u003e","description":"","filename":"FigureII.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8958644/v1/75b02c45d5456e5e8d1cfd83.jpg"},{"id":104397884,"identity":"dc5498b8-54d4-47d7-bfec-3b6df5e0c326","added_by":"auto","created_at":"2026-03-11 11:58:37","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":914615,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8958644/v1/8558b33f-ce1c-4c90-8a7b-1462d7b45a71.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"The Future of Shopping: AI Commerce","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eAs consumers rely on AI-based search, organic web traffic is reduced (Sommerfeld, 2025), impacting both SE Optimization (SEO) and SE Marketing (SEM). As Artificial intelligence (AI) advances, consumers are encountering new modes of interaction with retailers, such as chatbots using GenAI, which can be understood as Conversational AI (ConAI). OpenAI\u0026rsquo;s ChatGPT (Peters, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), Perplexity AI (Capoot and Rooney, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and Google\u0026rsquo;s Gemini (Mehta, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) are examples of ConAI that introduced features to enable product search, shopping and checkout which may substantially impact retail. As consequence, e-commerce and marketplaces may also suffer a reduction in traffic and conversion from traditional SEs.\u003c/p\u003e \u003cp\u003ePurchasing intention when using ConAI is correlated to its ease of use (Jan et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Lopes et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), usefulness, trust (Elsayad and Mamdouh, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Mari et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Selter et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Shahzad et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Silva et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), perceived risk, anthropomorphism, and intelligence (Mpinganjira et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Sohn et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Song et al., \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Enhancements to ConAI may increase purchase intention, leading to more users shopping and buying through it.\u003c/p\u003e \u003cp\u003eConcepts such as human digital twins (Barricelli et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Guo and Lv, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Wang et al., \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), digital twin of a customer (Gartner, 2023) and machine customers (Scheibenreif and Raskino, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) may evolve as new AI models are developed. ConAI would embed such concepts to better assist its users in searching and buying retail products.\u003c/p\u003e \u003cp\u003eBe it through advertising products from e-commerce and marketplaces or having merchants to sell directly through ConAI, LLMCs such as OpenAI may explore take rate fees as an additional revenue stream besides subscription models, as currently LLMCs are not profitable (Wiggers, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and need to monetize.\u003c/p\u003e \u003cp\u003eAll these social, technological, and economic signals point to a future where zero-click search for products and frictionless buying through AI shopping assistants becomes the norm for ConAI users. This article attempts to understand what happens to e-commerce and digital marketing as we know with the rise of ConAI based on LLMs. Through Speculative Design (SD) techniques (Dunne and Raby, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), more specifically SPM (Rhydderch, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), it is possible to forecast possible futures as well as extrapolate DPs using Design Fiction techniques (Bleecker, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2009\u003c/span\u003e) that depict such futures where advertising and commerce are greatly impacted. Hence, the following research question was formulated:\u003cdiv class=\"BlockQuote\"\u003e\u003cp\u003e \u003cem\u003eRQ.\u003c/em\u003e Can ConAI replace e-Commerce?\u003c/p\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eAs such, this article contributes to online retailing and retail management practice by offering an overview of future possibilities of ConAI in the industry.\u003c/p\u003e"},{"header":"2. Methodology","content":"\u003cp\u003eAccording to Dunne and Raby (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), SD is a way to explore complex problems and possible futures through imagination and questioning the \u003cem\u003estatus quo\u003c/em\u003e. This approach to design flourishes through imagination, seeking to shed new light on so-called wicked problems, foster dialogue and debate about different ways of living, and spark and nurture people\u0026rsquo;s creative thinking (Dunne and Raby, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eSD serves to materialize the multitude of potential futures our world could inhabit. While it is traditionally understood that the present is determined by historical events, it is equally valid to consider how it may be shaped by projections of the future\u0026mdash;by our collective hopes, ambitions, and imagined possibilities (Dunne and Raby, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThrough SD techniques, we forecast future scenarios with the purpose of using these ideas as tools to better understand the present and to serve as a discussion point for future analysis.\u003c/p\u003e \u003cp\u003eTo enable the forecast of such future scenarios, we investigate signals from different sources, such as academic papers and news reports.\u003c/p\u003e \u003cp\u003eFor academic papers, a literature review was done using different search strings within the Scopus database. To understand signals on GenAI and e-Commerce, a search combining \u0026ldquo;Artificial Intelligence\u0026rdquo; AND \u0026ldquo;e-Commerce\u0026rdquo; was used. To understand purchase intention when using GenAI, a search string combining \u0026ldquo;purchase intention\u0026rdquo; AND \u0026ldquo;artificial intelligence\u0026rdquo; was used. And to understand the impacts of GenAI personalization, a string with \u0026ldquo;digital twin\u0026rdquo; AND \u0026ldquo;human\u0026rdquo;. Results were further refined using filters for papers in the timeline of 2019 to 2025 and an exclusion criterion based on relevance to the theme of this research.\u003c/p\u003e \u003cp\u003eWe also collected news and LLMCs\u0026rsquo; announcements on advancements of generative AI solutions into retail, especially e-Commerce and digital marketplace practices, as well as news and reports on the impact of GenAI searching into SEO/SEM.\u003c/p\u003e \u003cp\u003eBased on this collection of signals, we extrapolated possible future scenarios using SD techniques.\u003c/p\u003e \u003cp\u003eThe scenarios are built on a what-if approach (Dunne and Raby, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) under a SPM (Rhydderch, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). By gathering and clustering signals, we can plot the x and y axes on the SPM and use that planning technique to come up with what-ifs under the parameters established for each of the four spaces formed by the intersection of both axes. In the possible futures section, we will further explain the context sustaining our speculative scenarios.\u003c/p\u003e \u003cp\u003eUnder each of the four spaces, it is possible to create several different speculative scenarios. For the purpose of this research, we\u0026rsquo;ll briefly name all the scenarios that were speculated for the SPM but focus only on the ones that were deemed to be the most probable and preferable, as per the Probable, Plausible, Possible, Preferable - PPPP standard (Dunne and Raby, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTo materialize those chosen scenarios, we\u0026rsquo;ll combine the two main speculations (probable and preferable) with a Design Fiction approach: DPs \u0026ndash; fictional pieces inserted into that future scenario (Bleecker, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2009\u003c/span\u003e).\u003c/p\u003e"},{"header":"3. Signals Collected","content":"\u003cp\u003eThis section collects recent news and academic papers representing relevant signals to this research.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eI. GenAI in e-Commerce\u003c/p\u003e\n\u003cp\u003eResearch on AI in retail focuses primarily on recommender systems, especially for e-Commerce. Guha \u003cem\u003eet al\u003c/em\u003e. (2021) indicates that AI helps by making omnichannel and mobile shopping more profitable, notably by sharpening personalized recommendations. Advanced AI algorithms enable e-commerce platforms to interpret external data, learn from it, and improve recommendation quality through flexible adaptation (Bawack \u003cem\u003eet al\u003c/em\u003e., 2022). Personalized recommendation systems have become a vital tool for online retailers, enhancing user experience through tailored product suggestions based on individual preferences and behaviours (Dharwadkar and Kannadaguli, 2024).\u003c/p\u003e\n\u003cp\u003eGenAI offers e-commerce capabilities to synthesize a large amount of customer-provided data (prompts), supporting consumers in their shopping journeys with new retailing experiences (Al-Shaikh \u003cem\u003eet al\u003c/em\u003e., 2024). GenAI capabilities span multiple areas: product description and content generation, product image generation, customer support chatbots, personalization and recommendation, product review analysis, and marketing and advertising (Ghaffari, 2024).\u003c/p\u003e\n\u003cp\u003eConAI in e-commerce naturally answers shopping queries, recommends relevant products, finds product descriptions, and links them to platform products. ChatGPT-driven chatbots provide personalized, interactive experiences improving efficiency and value proposition, offering higher responsiveness and accuracy, reducing consumer effort and time while providing product knowledge (Yue and Basha, 2024).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003ePantano and Pizzi (2020) showed an increasing number of chatbot patents in the past 20 years, suggesting the use of new conversational agents based on natural language might lead to better interactions between consumers and retailers.\u003c/p\u003e\n\u003cp\u003eThe intersection between AI and customer experience in retailing shows chatbots and voicebots represent most researched AI technologies. Studies highlight that chatbots maintain customer interest through convenience, ubiquity, interactivity, technology readiness, ease of use, satisfaction, and trust (Peruchini \u003cem\u003eet al\u003c/em\u003e., 2024).\u003c/p\u003e\n\u003cp\u003eII. GenAI replacing SEs\u003c/p\u003e\n\u003cp\u003eInitial research suggested ChatGPT-like technology would not replace commercial SEs like Google and Bing, expecting LLMs to become SE components serving as additional query options (Strzelecki, 2023).\u003c/p\u003e\n\u003cp\u003eSince OpenAI\u0026apos;s LLM integration into Microsoft Bing and Google\u0026apos;s AI mode introduction, widespread speculation emerged about ConAI becoming \u0026quot;answer engines\u0026quot; replacing traditional SEs. Conversely, others argue SEs aren\u0026apos;t facing obsolescence, expecting search and chat functionalities to coexist harmoniously (Barnard and Artz, 2023).\u003c/p\u003e\n\u003cp\u003eChatGPT is used by students and industry as a SE to enhance work and research scope, quality, and creativity with daily or weekly frequency (MacLachlan \u003cem\u003eet al.\u003c/em\u003e, 2024). Such ConAI use increases despite research showing users trust ChatGPT less than Google and Wikipedia (Jung \u003cem\u003eet al\u003c/em\u003e., 2024).\u003c/p\u003e\n\u003cp\u003eBarnard and Artz (2023) note that the future of search marketing will depend on how search and answer engines understand brand identity, evaluate brand credibility, and deliver and prioritize brand content.\u003c/p\u003e\n\u003col start=\"3\"\u003e\n \u003cli\u003eThe Fall of SEO\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eWith ConAI replacing traditional SEs, there are impacts to search marketing and online advertising (Gartner, 2023). SEO, the traditional set of techniques to increase visibility in organic SE results, is also being impacted, with metrics such as click-through rate (CTR) metrics reducing drastically (Sommerfeld , 2025).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRecent survey from Bain \u0026amp; Company (Sommerfeld \u003cem\u003eet al.\u003c/em\u003e, 2025) finds that about 80% of consumers now rely on GenAI overviews delivered by traditional SEs (for example, Google and Bing) at least 40% of the time, reducing organic web traffic by an estimated 15% to 25%. As a user may perform no clicks after reading an AI overview on Google or after a reply from ChatGPT, this type of search can be called a zero-click search, which impacts both SEO and SEM.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eA study conducted by Seer Interactive (Haigler, 2024) shows that organic CTR declined about 70% when an AI overview was present on the SE results page. A similar study analysed by Ahrefs (Law and Guan, 2025) shows 34.5% lower average CTR for the top-ranking page, while Amsive (Guevara, 2025) points to a CTR decline of -15.49% on average.\u003c/p\u003e\n\u003cp\u003eEven though these studies show different decline rates for CTR, all of them point to a reduction in CTR. CTR is one of the most important metrics in both SEO and SEM, with the earlier improving ranking of a page on SE results and the later improving ranking amongst sponsored links. SEO services are forecasted to reach US$ 106,9 billion in 2025 (Research and Markets, 2025) while SEM spending is forecasted at US$351.55 billion in 2025 (Statista, 2025). Such market sizes show that the growth of AI assistants based on LLM may have while reducing CTR metrics.\u003c/p\u003e\n\u003cp\u003eCorroborating these studies, Gartner (2023) expects that by 2026 traditional SE volume will drop 25%, as SEM will lose market share to AI chatbots and other virtual agents.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBain (Sommerfeld \u003cem\u003eet al\u003c/em\u003e., 2025) comments on the rise of LLMs for search pointing that roughly 40% to 70% of LLM users use the platforms to conduct research and summarize information (68%), understand the latest news and weather (48%), and ask for shopping recommendations (42%).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAdobe Analytics data shows that consumers are already using ConAI to get product recommendations (Pandya, 2025). This data shows that ConAI traffic to United States retail sites has increased by 1.300% between November and December 2024 and the same period in 2023. This trend seems to have persistence, as traffic increased by 1.200% between July 2024 and February 2025.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAdditional ConAI solutions are expected to come to market from companies such as Apple, Samsung, Amazon and Meta, as well as startups and will be embedded in smartphones and other devices by default, providing further search substitutions, personalization and ease of use (Gartner, 2023).\u003c/p\u003e\n\u003cp\u003eAs ConAI users may interact only with the AI itself, with no clicking through additional pages, further adoption of ConAI may impact not only web page traffic, but e-Commerce as well.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIII. GenAI as shopping assistant\u003c/p\u003e\n\u003cp\u003eMatos \u003cem\u003eet al\u003c/em\u003e. (2025) points that as GenAI advances, consumers are encountering new modes of interaction with retailers, such as ConAI. Gupta and Mukherjee (2024) suggest that GenAI search tools can help make the search process simple, convenient, less effortful, and less time consuming, providing better search experiences and improving the chances of arriving at the best purchase decision, showing that GenAI can potentially disrupt consumer\u0026rsquo;s online search behaviour. Huh and Kim (2024) also indicate that consumers perceive ConAI as and efficient and convenient decision-aid tool for smart shopping, highlighting utilitarian features such as speed and convenience for customers as well as hedonic values as an emotional companion for shopping.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRecent examples of customer facing ConAI include Alexa, Amazon\u0026rsquo;s voice assistant which interacts with customer primarily via voice and is increasingly used for online shopping (Guha \u003cem\u003eet al.\u003c/em\u003e, 2021).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eLately, as ConAI gains traction with users, LLMCs introduce new features include product search and recommendations as well as shopping and checkout. \u0026nbsp;Perplexity AI introduced e-Commerce features for retail shopping, including the ability to check out right on their ConAI app or website (Perplexity, 2024). These features allow users to take a photo of an item to search for products as well as to see traditional product cards when users search through shopping questions. Product recommendation is not sponsored, with expectation of unbiased search results.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOpen AI has also gradually introduced features to enable e-Commerce, previewing new ChatGPT features for users to do research around product showing product cards with images, prices and ratings (Peters, 2025).\u003c/p\u003e\n\u003cp\u003eAs hyperscalers such as Google and Amazon are developing their own LLMs and ConAI applications, e-Commerce-like capabilities are also being explored by those.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eGoogle has announced agentic checkout and shopping features to be embedded on its SE. New shopping features are also being introduced to Google\u0026rsquo;s AI Mode, allowing users to engage in a conversation with considerations on product reviews, prices, colour options and availability. Google\u0026rsquo;s AI Mode will enable virtual try-ons for clothes and agentic checkout to track prices for users (Rincon, 2025).\u003c/p\u003e\n\u003cp\u003eAmazon has introduced ConAI-features to its platform through an expert shopping assistant called Rufus (Mehta and Chilimbi, 2024) to answers users\u0026rsquo; questions on shopping needs, products and comparisons and make recommendations.\u003c/p\u003e\n\u003cp\u003eAmazon has also launched a feature called \u0026ldquo;Buy for Me\u0026rdquo; (Amazon, 2025), leveraged through agentic AI capabilities, allowing customers to purchase products directly from other websites\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIV. Purchase Intention through ConAI\u003c/p\u003e\n\u003cp\u003eJan \u003cem\u003eet al\u003c/em\u003e. (2023) found four major reasons for using ConAI for retailing. These are\u003c/p\u003e\n\u003cp\u003eperceived usefulness, perceived ease of use, perceived trendiness, and perceived informativeness, which have significant positive effects on attitudes toward the usage of ConAI. Jan \u003cem\u003eet al\u003c/em\u003e. (2023) also discusses three major reasons against ConAI for shopping. These are the usage barrier, functional risk barrier, and intrusiveness barrier for ConAI.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSong \u003cem\u003eet al.\u0026nbsp;\u003c/em\u003e(2022) found that perceived communication quality and privacy risk mediate the effect of ConAI on adoption intention. While AI personalized services increase convenience while increasing users\u0026rsquo; privacy concerns, Song \u003cem\u003eet al.\u0026nbsp;\u003c/em\u003e (2022) also perceives that consumers\u0026rsquo; willingness to accept ConAI is enhanced if they perceive lower privacy risks during the interaction.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eHowever, Sohn \u003cem\u003eet al.\u003c/em\u003e (2025) find that consumers express lesser privacy concerns in the presence of AI (vs. human) service agents. This effect stems from consumers\u0026rsquo; perceptions that AI agents have less power than human agents.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eElSayad and Mamdouh (2024) identifies trust as critical in customer interactions with smart technologies, particularly in e-commerce where face-to-face interaction is absent. Customers rely on trust in platform reliability, security and integrity, shaping perceptions and decisions throughout online shopping. Trust fundamentally impacts perceived service value and purchase decisions. Customer technology readiness influences trust in AI-based retail services through motivators (optimism and perceived innovativeness) and inhibitors (discomfort and insecurity). Technology readiness significantly predicts perceived trust, subsequently influencing perceived usefulness and purchase intentions.\u003c/p\u003e\n\u003cp\u003eSilva \u003cem\u003eet al. (\u003c/em\u003e2023) also indicate that user trust significantly predicts behavioural intention toward ConAI, while also reducing perceived risk and enhancing flow experience.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eHuh and Kim (2025) have found that utilitarian and technological value are more important than hedonic and social value. As such speed and convenience are central to retailing through ConAI.\u003c/p\u003e\n\u003cp\u003eOn the hand, Selter \u003cem\u003eet al\u003c/em\u003e. (2023) found no significant results for the utilitarian and hedonic product types, but rather a decreasing usage intention for automated shopping processes. In addition, Selter \u003cem\u003eet al\u003c/em\u003e. (2023) found that trust and behavioural control mediate the effect of automation on usage intention, and this effect is further moderated by inherent novelty seeking.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eShahzad \u003cem\u003eet al.\u0026nbsp;\u003c/em\u003e(2024) found that trust is tied to customers\u0026rsquo; perceptions of the privacy and security of their data during interactions with the ConAI.\u003c/p\u003e\n\u003cp\u003eLopes \u003cem\u003eet al\u003c/em\u003e. (2024) shows AI in online retailing significantly alters consumer buying behaviour through personalized recommendations, simplifying decision-making and increasing purchase likelihood. Digital shopping behaviour is increasingly influenced by psychosocial factors shaping purchase intention: subjective norms (community expectations influencing AI attitudes), faith (trust in AI), consciousness (ethical awareness) and perceived control (belief in controlling AI). When these factors favourably impact AI-enabled ease of use, purchase intention is perceived favourably.\u003c/p\u003e\n\u003cp\u003eMpinganjira \u003cem\u003eet al\u003c/em\u003e. (2023 found that the continued use of chatbots is primarily driven by experiential utilitarian value, followed by hedonic value. However, perceived ConAI risk negatively influences customer experience value in retailing due to concerns about exposing sensitive personal information to unauthorized users.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMari \u003cem\u003eet al\u003c/em\u003e. (2024) determines correlation between trust, usefulness and perceived empathy in ConAI and its use as decision aid assistant, delegated agent or trusted product recommender.\u003c/p\u003e\n\u003cp\u003eOpposing such studies, Aiolfi (2023) found that perceived privacy risk, innovativeness and social attraction have been found to not significantly impact attitudes towards ConAI, in particular smart speakers. One possible justification might come from the Privacy Paradox: although people say they care about their privacy and are not willing to share their data, actually they give their private information in exchange for small benefits or convenience.\u003c/p\u003e\n\u003cp\u003eV. Digital Twin of a Customer\u003c/p\u003e\n\u003cp\u003eDigital twins are computer-based models simulating physical entities\u0026mdash;objects, processes, humans or human features (Barricelli \u003cem\u003eet al\u003c/em\u003e., 2019). Wang (2024) explores human digital twins that reflect multi-dimensional human information\u0026mdash;physical, psychological, and cognitive data\u0026mdash;enabling modelling and simulation of mental workload, interactions and trust to replicate human decision-making roles.\u003c/p\u003e\n\u003cp\u003eIn retail, Guo and Lv (2022) examines digital twins for user-product interaction in offline and online sales, enabling shoppers to experience goods through virtual reality using digital twins of users and products. Gartner (2023) suggests that digital twins of a customer will allow for simulation and anticipation of customer behaviour. Yang \u003cem\u003eet al\u003c/em\u003e. (2025) shows that digital twins have been used in retail to provide virtual replicas of products, service systems, shopping environments and customers interactions allowing consumers to experience products in real-time and reduce product misfit and return losses.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eScheibenreif and Raskino (2023) predicted machine customers would outperform humans at information searching, deal negotiation, purchasing, and other buying processes, only requiring trust in both technology and privacy for larger consumer adoption.\u003c/p\u003e\n\u003cp\u003eVI. LLMCs Monetization Problem\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eLLMCs have raised significant resources and have large valuation figures, with Open AI funding at US$ 40 billion, Perplexity AI at US$ 14 billion and Anthropic at US$ 3.5 billion. Despite such figures and growing monthly users, it seems these companies are yet to become profitable (Wiggers, 2025). For instance, ChatGPT\u0026rsquo;s Open AI reported US$ 5 billion loss in 2024, despite US$ 5.5 billion in revenue from subscriptions (Field, 2025).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn a not-so-distant past, big tech companies such as Google and Facebook started monetizing by capturing advertising revenue (Curran \u003cem\u003eet al\u003c/em\u003e., 2011). Google generates revenue by monetizing its SE services through ads services, where advertisers pay money to have their listing ranked higher in search results (KRITZINGER, 2013). Most SEs offer pay-per-click practices to generate revenue (Seymour \u003cem\u003eet al\u003c/em\u003e., 2011). Facebook offers a model to advertise and engage with consumers (Curran \u003cem\u003eet al.\u003c/em\u003e, 2011). Instagram also relies on advertising to monetize (Burgess \u003cem\u003eet al.\u003c/em\u003e, 2024).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIt is not farfetched to imagine that LLMCs will try and capture advertising revenue, besides monetizing through a subscription model.\u003c/p\u003e"},{"header":"4. Possible Futures: 2x2 SPM","content":"\u003cp\u003eCollected signals perceived in social, economic, and technological spaces were used to support defining axes in the 2x2 SPM.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIt is clear that LLMCs need to increase their revenue and may choose to do so through a stream different from the existing subscription model. LLMCs are investing in capabilities for product recommendation and one-click checkout, which may lead to introduction ads in their ConAI. LLMCs may choose to sell merchants\u0026rsquo; products, similarly to marketplaces. These choices are reflected in the SPM.\u003c/p\u003e\n\u003cp\u003eWe define the vertical axis of SPM by poising the question \u0026ldquo;How will LLMCs monetize besides subscriptions?\u0026rdquo; as seen in Figure I. As variables in this axis, answering the questions, we propose the choice of \u0026ldquo;Selling\u0026rdquo; and \u0026ldquo;Advertising\u0026rdquo;.\u003c/p\u003e\n\u003cp\u003eEvaluating customer purchase intention through ConAI, easy usage and trust are key factors for users to shop through an AI. As we consider easy usage a matter of time and investment, trust is the real psychological factor that may decide whether ConAI will succeed in either showing product ads or selling merchant products. As we understand trust may be a concept difficult to grasp as a product feature that a LLM company can embed on a ConAI, we consider user data a better simulacrum for trust. So, we chose to poise the question \u0026ldquo;Do people trust their own data to AI assistants?\u0026rdquo; On the horizontal axis of the SPM, as seen in Figure I.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAdd Figure I here\u003c/p\u003e\n\u003cp\u003eWe speculate four speculative scenarios, as seen on Figure I:\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSpeculative Scenario 1:\u0026nbsp;\u003c/strong\u003e\u003cem\u003eLLMCs monetize through selling and consumers trust their data uncritically to ConAI\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eLLMCs invest in shopping and checkout capabilities embedded in their ConAI sites and apps. As consumers continue to provide personal data to ConAI, these solutions end up becoming digital personal shoppers through agentic AI features and evolve into human digital twins as LLMs become capable of representing virtually any consumer. LLMCs introduce API communication to facilitate any vendor or merchant to add product details such as description, specs, photos, videos, price and stock availability so ConAI can better recommend products to their users. LLMCs charge a take rate fee for every merchant sale to consumers through ConAI sites or apps. As buying through ConAI becomes the standard, traditional retail e-commerce and marketplace platforms end-up as obsolete. In this scenario, we speculate that AI commerce, where consumers purchase goods and services through ConAI, becomes the standard for any online purchasing.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSpeculative Scenario 2:\u0026nbsp;\u003c/strong\u003e\u003cem\u003eLLMCs monetize through selling, but consumers are sceptical about trusting their data to ConAI\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eLLMCs monetize selling goods and services for other vendors and merchants through ConAI sites and apps. In this scenario consumers are sceptical to trust their personal data to a ConAI, not willingly providing demographic, behavioural, and transactional data. With this limitation, product search and recommendation are not user-oriented, but rather generic, with no personalization. As perceived ease of use and trust are suboptimal in this scenario, purchase intention is low (Lopes \u003cem\u003eet al\u003c/em\u003e., 2024). Therefore, sales figures through ConAI are also low. We speculate that in this scenario, only geeks and early adopters end up using ConAI for shopping and checkout. We speculate that LLMCs will implement other strategies to increase perceived trust, such as brand partnerships to increase sales by offering discounts for group sales in bulk, as well as loyalty programs for their users, offering cashback. However, we speculate that such tactics will fail and LLMCs will end up not breaking even, which would force them to sell to hyperscalers to avoid bankruptcy in the long run.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSpeculative Scenario 3:\u003c/strong\u003e \u003cem\u003eLLMCs monetize through advertising and consumers trust their data uncritically to ConAI\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eLLMCs chose to increase their revenue by pursuing advertising budgets from other companies while consumers feed ConAI with personal data uncritically. With such data, ConAI solutions can offer hyperpersonalized product recommendations, which deliver greater CTR and conversion to e-commerce and marketplace sites that advertise through the ConAI. This evolution renders SEO and SEM completely obsolete. In addition to ads revenue, LLMCs continue to collect revenue from ConAI subscriptions. Following trends observed in other subscription models, LLMCs introduce tiers to ConAI subscription, including an ads free premium tier.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSpeculative Scenario 4:\u0026nbsp;\u003c/strong\u003e\u003cem\u003eLLMCs monetize through advertising, but consumers are sceptical about trusting their data to ConAI\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eLLMCs chose advertising as new business model, however consumers do not trust their personal data to ConAI. This leads ConAI to provide unbiased product recommendations to its users. We speculate that this will make ConAI accepted but not used to a degree where SEs may completely disappear. As such, SEO and SEM continue to evolve, further developing features such as AI Overview and AI Mode. With continued competition with traditional SEs, now fully embracing AI, LLMCs do not break even and end up acquired by hyperscalers, as LLMs replace traditional search indexers, becoming more of a relevant tool than a complete solution.\u003c/p\u003e"},{"header":"5.\tDiegetic Prototypes","content":"\u003cp\u003ea. Most Probable Future (Speculative Scenario 1)\u003c/p\u003e\n\u003cp\u003eAs the most probable future, we pictured a world where AI Commerce is present in everyone\u0026rsquo;s life one way or the other. In that world, a Podcast called \u0026ldquo;Checkout\u0026rdquo; becomes popular. Here is a transcript of one the show\u0026rsquo;s episodes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003e\u0026ldquo;Checkout Podcast Season 3 Episode 11 - The Gallery\u0026rdquo;\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eOpening:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;You\u0026rsquo;re listening to Checkout \u0026mdash; the number 1 retail-tech podcast for buyers and sellers.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;Today, we are checking out the stories behind a new luxury market that quietly emerged while we were all still online shopping. And joining me is our beloved Vander.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: Thanks, Bela. Out of Home is no longer just a media strategy term; it\u0026rsquo;s become the latest hype among high-ticket consumers. But before we dive deeper, we\u0026rsquo;ve got a special guest with us today, right?\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: That\u0026rsquo;s right! Our dearest digital influencer, Miquela, a.k.a. \u0026ldquo;@lilmiquela\u0026rdquo;, is here with a message from OpenBrain.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eMiquela\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: Hi Checkoutters! OpenBrain has an amazing update \u0026ndash; I can be your digital twin for FREE! Subscribers, just login and select \u0026quot;MyQuela.\u0026quot; Customize your shopping preferences and create a MyQuela that looks like you. I automate recurring purchases, find the best deals, and send you releases matching your style. Browse OpenBrain while you listen \u0026ndash; I handle all payments. Non-subscriber, stick around for a gift!\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;Thanks, Miquela. I already have my digital twin with MyQuela, and what a joy! Everything arrives without me even having to think about it. So practical, so fast.\u0026nbsp;\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;Not to mention the savings! I\u0026rsquo;ve been a subscriber since MyQuela launched the twin feature back in 2030, and I noticed the difference in my budget right away! It finds exactly what I want, at the best price and highest quality.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;Exactly! And look, while you and I are using our digital personas to speed up shopping, there\u0026rsquo;s a crowd who wants to live the full in-person shopping experience. And that\u0026rsquo;s who we\u0026rsquo;re talking about today.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: That\u0026rsquo;s right. Today, we\u0026rsquo;re talking about the new offline trend in the luxury market. The Gallery is a new Shopping World with stunning landscapes, top-notch hospitality, diverse brands, and eclectic cuisine.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: Shopping World? Don\u0026rsquo;t you mean shopping malls?\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: laughs. No, Bela. I really do mean World. Here\u0026rsquo;s the deal: since 2025, even before digital twins and AI shopping tools became mainstream, luxury brands had already noticed two big trends: a strong focus on longevity and the need to sell experiences more than just products. That\u0026rsquo;s how we started seeing initiatives like Prada Caff\u0026egrave;, Bar Luce, Gucci Osteria, The Tiffany Blue Box Caf\u0026eacute;, the Louis Vuitton restaurant...\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;I remember when all that started! Now we have that luxury Ralph Lauren hospital in S\u0026atilde;o Paulo and that Tommy Hilfiger Speed Track! I\u0026rsquo;ve always wanted to try it.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;Oh, it is amazing. I celebrated my birthday there last month, and it was awesome. And get this: Yang Huiyan picked up on this trend and launched the first Shopping World in Shenzhen in 2036. Imagine a theme park, but for adults who love luxury brands and have the time to do in-person shopping.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: Yang Huiyan, you mean the Real Estate Chinese Billionaire?\u0026nbsp;\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: Precisely. She had the means and was deeply involved with her audience to start this new business.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003cem\u003eI mean, it does make sense. And I am assuming that the name for her new enterprise is no coincidence. The Gallery. I bet it has something to do with her background in the arts.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: You are absolutely right. She is a Bachelor of Arts/Science by Ohio State University, and that influenced her decision for the name and the whole business model. Her Western experience gave her a broader understanding of appealing to both Westerners and Easterners. The artistic philosophy behind The Gallery\u0026rsquo;s success: every product has to be fine-art-like, and every purchase should be special.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: So, you are not just taking a bag home, you are purchasing a feeling. Brilliant! But, can you imagine the ticket prices?! *laughs*\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: That was another genius move: no tickets \u0026mdash; invitation only. Total exclusivity. For The Gallery\u0026rsquo;s first edition, she partnered with 45 brands that had to design unique experiences specifically for that World. Patek Philippe, for example, built a giant watchmaking workshop where visitors could create their own watch straps. A brand that values tradition and longevity created a space where parents and kids could share a memory that would last forever and keep perfect time too.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: That\u0026rsquo;s incredible! It makes sense she would target only the richest. Who else got the time to do in-person shopping in this economy?\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: Right? The current scenario is: we\u0026rsquo;ve still got folks shopping on the web using their favourite LLM (the late AI adopters). Then the growing crowd using human digital twins. And finally, there\u0026rsquo;s this luxury crowd using their time and money traveling abroad in search of offline shopping experiences.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: Funny how you didn\u0026rsquo;t even mention the old SEs *laughs*. I don\u0026rsquo;t think I\u0026rsquo;ve opened one for shopping since 2030. Do any still exist?\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;There are still a few survivors on life support. Almost no one uses a browser for searching \u0026mdash; let alone shopping \u0026mdash; especially with so many free conversational models available.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: And I also haven\u0026rsquo;t seen any in-person shopping/market/mall. There are some small garage sales, but since no one has the time to go places and do in-person purchases, we\u0026rsquo;ve all been online shopping with drone-dropped deliveries.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;Right? It\u0026rsquo;s not about money anymore. No one has the time to drive far away just to pick some apples and oranges. Let AmazonFresh handle that.\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e: Amazon, if you are listening, we\u0026rsquo;d love to partner with you *laughs*. I still think it\u0026rsquo;s crazy that those giant tech companies from the past either died or had to completely reinvent themselves.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;Totally. Last year, 2039, I saw the last major bankruptcy. Now, the old SEO teams are trying to grab a slice of LLM product recommendation space and cramming in Ads wherever they can. But for premium subscribers, it\u0026rsquo;s all fully personalized, zero ads. It\u0026rsquo;s amazing.\u0026nbsp;\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;Smart were the tech companies that pivoted early. Besides the premium subscriptions; they also earn a fee on every purchase completed through the LLM. Is that true?\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;It is. For users without purchase profiles, tech companies make money through ads. For those who use digital twins or premium shopping features like automation, they win a fee for each purchase.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003cem\u003eJust thinking about all this makes me want to visit that Shopping World. Maybe play soccer with Emirates-sponsored players? *laughs*\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;While we\u0026rsquo;re here automating daily shopping and leaving the mall runs to our digital twins, some folks have time to spend days immersed in those incredible experiences.\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;Who knows, maybe by 2050 I\u0026rsquo;ll be one of them? Let\u0026rsquo;s go to Shopping World Brazil edition?\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;Let\u0026rsquo;s go! Maybe we\u0026rsquo;ll land a partnership with Yang Huiyan and score an invite. Until then, digital twins it is.\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBela:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;For sure! As promised, for those of you who stayed till the end, go to OpenBrain use our promo code for the premium plan: \u003cstrong\u003eCHECKOUT30\u003c/strong\u003e. Thanks, Miquela, OpenBrain!\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eVander:\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e\u0026nbsp;Thanks, everyone. We are checking out.\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e(DP - Developed by the authors)\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eb.\u0026nbsp;\u003c/em\u003ePreferable Future\u003c/p\u003e\n\u003cp\u003eAs a preferable scenario, we pictured a world where some users approach data trust more critically, with a better balance between ad usage for AI Commerce monetization.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOn that future, people can choose how much they want to invest in their own privacy, however more control means a more costly option for AI commercial usage, while less control means lower prices for AI usage. This scenario is closer to the intersection point of the two axes in the SPM presented in Figure I.\u003c/p\u003e\n\u003cp\u003eAdd Figure II here\u003c/p\u003e\n\u003cp\u003eUnder the preferable scenario, as seen in the UI on Figure II, data use is limited. For instance, there is no Digital Twin available for Free plan, with data largely taken from users to enhance ad segmentation, making that plan still profitable for AI companies. Data usage and storage is still limited according to applicable laws and regulations, which users cannot consent to under the free plan and thus makes it impossible for that option to have a digital twin as this would require personal data that cannot be given without consent.\u003c/p\u003e\n\u003cp\u003eThe second plan and third plan differ precisely on how much data can be taken from users and how much they can control its use for all AI actions, though these plans are more costly.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e"},{"header":"6. Conclusion","content":"\u003cp\u003eIn this article we collected signals that were used to plot a SPM and create DP using SD and Design Fiction techniques.\u003c/p\u003e \u003cp\u003eBased on the SPM, we speculated possible futures where LLMCs monetize through advertising and/or selling merchant products directly on ConAI. We speculate whether users would trust their personal data to ConAI uncritically or sceptically.\u003c/p\u003e \u003cp\u003eWe further explore possible futures, understanding that the most probable future has AI commerce dominating, with ConAI replacing SEs as well as traditional retail e-commerce and marketplaces. We also speculate on a preferable future, where users may choose different subscription tiers on ConAI solutions to better control their personal data as well as exposition to ads.\u003c/p\u003e \u003cp\u003eThough this speculative exercise aims to be comprehensive when it comes to ConAI, SEs, LLMCs, human digital twins and psychological factors influencing consumers, it has been limited towards social commerce trends and implications for the labour market and other social impacts. Also, this article does not explore and the future of shopping on physical stores or the impact of AI on physical stores, although DP speculated in this paper reaffirm the domination of AI commerce in retailing and hypothesize a future where all physical retail stores become a luxury. Further exercises may encompass discussion on these areas.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eI.G. contributed with the methodology on speculative design while V.R. gathered the signals from academic and non-academic sources. Both authors worked on the building one the possible and probable futures. V.R. created the 2x2 matriz. I.G. created the design fiction pieces. Both authors reviewed the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThe authors acknowledge methodology guidance on speculative design techniques from Prof. Diogo Cortiz, PhD, theoretical support on artificial intelligence from Prof. Dora Kauffman, PhD as well as support from Bruno Johnson, MS.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAiolfi S. (2023), \u0026quot;How shopping habits change with artificial intelligence: smartspeakers\u0026apos; usage intention\u0026quot;. \u003cem\u003eInternational Journal of Retail \u0026amp; Distribution Management\u003c/em\u003e, Vol. 51 No. 9-10 pp. 1288\u0026ndash;1312, doi: doi.org/10.1108/IJRDM-11-2022-0441\u003c/li\u003e\n\u003cli\u003eAl-Shaikh, M.S., Al-Mousa, M.R., Al-Ababneh, H.A., Ali, M.W., Alkaawneh, S.M., Barhoush, F.M.S. and Binsaddig, R. (2024), \u0026ldquo;The challenges of using generative AI applications in electronic commerce\u0026rdquo;, in IEEE (Ed.), \u003cem\u003eProceedings of the 25th International Arab Conference on Information Technology (ACIT 2024), \u003c/em\u003eIEEE, Piscataway, NJ, pp.1\u0026ndash;5, doi: doi.org/10.1109/acit62805.2024.10876935\u003c/li\u003e\n\u003cli\u003eAmazon (2025), \u0026ldquo;Amazon Shopping App: \u0026lsquo;Buy for Me\u0026rsquo; feature enables purchases from other brands\u0026rsquo; sites\u0026rdquo;, available at: https://www.aboutamazon.com/news/retail/amazon-shopping-app-buy-for-me-brands (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003eBarnard, J. and Artz, M. (2023), \u0026ldquo;Search marketing in the age of AI: understanding the marketing implications of search, assistive and answer engines\u0026rdquo;, \u003cem\u003eJournal of Digital \u0026amp; Social Media Marketing\u003c/em\u003e, Vol. 11 No. 3, p.244, doi: doi.org/10.69554/zvmd6442\u003c/li\u003e\n\u003cli\u003eBarricelli, B., Casiraghi, E. and Fogli, D. (2019), \u0026ldquo;A survey on digital twin: definitions, characteristics, applications, and design implications\u0026rdquo;, \u003cem\u003eIEEE Access\u003c/em\u003e, Vol. 7, pp.167653\u0026ndash;167671, doi: doi.org/10.1109/ACCESS.2019.2953499\u003c/li\u003e\n\u003cli\u003eBawack, R.E., Wamba, S.F., Carillo, K.D.A. and Akter, S. (2022), \u0026ldquo;Artificial intelligence in e-commerce: a bibliometric study and literature review\u0026rdquo;, \u003cem\u003eElectronic Markets\u003c/em\u003e, Vol. 32, pp.297\u0026ndash;338, doi: doi.org/10.1007/s12525-022-00537-z\u003c/li\u003e\n\u003cli\u003eBleecker, J. (2009), \u003cem\u003eDesign Fiction: a short essay on design, science, fact and fiction\u003c/em\u003e, Near Future Laboratory, Los Angeles, CA.\u003c/li\u003e\n\u003cli\u003eBurgess, J., Carah, N., Angus, D., Obeid, A. and Andrejevic, M. (2024), \u0026ldquo;Why am I seeing this ad? The affordances and limits of automated user-level explanation in Meta\u0026rsquo;s advertising system\u0026rdquo;, \u003cem\u003eNew Media \u0026amp; Society\u003c/em\u003e, Vol. 26 No. 9, pp.5130\u0026ndash;5149, doi: doi.org/10.1177/14614448241251796\u003c/li\u003e\n\u003cli\u003eCapoot, A. and Rooney, K. (2025), \u0026ldquo;OpenAI hits $10 billion in annualized revenue fueled by ChatGPT growth\u0026rdquo;, available at: https://www.cnbc.com/2025/06/09/openai-hits-10-billion-in-annualized-revenue-fueled-by-chatgpt-growth.html (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003eCurran, K., Graham, S. and Temple, C. (2011), \u0026ldquo;Advertising on Facebook\u0026rdquo;, \u003cem\u003eInternational Journal of E-Business Development\u003c/em\u003e, Vol. 1.\u003c/li\u003e\n\u003cli\u003eDharwadkar, V., R, V., S, M., M G, J. and Kannadaguli, P. (2024), \u0026ldquo;Smart cart: revolutionizing e-commerce in India with AI-powered personalized product recommendations overview\u0026rdquo;, in IEEE (Ed.), \u003cem\u003eProceedings of the 5th International Conference on Circuits, Control, Communication and Computing (I4C 2024), \u003c/em\u003eIEEE, Piscataway, NJ, pp.87\u0026ndash;92, doi: doi.org/10.1109/i4c62240.2024.10748521\u003c/li\u003e\n\u003cli\u003eDunne, A. and Raby, F. (2013), \u003cem\u003eSpeculative Everything: design, fiction, and social dreaming\u003c/em\u003e, MIT Press, Cambridge, MA.\u003c/li\u003e\n\u003cli\u003eElsayad, G. and Mamdouh, H. (2024), \u0026ldquo;Are young adult consumers ready to be intelligent shoppers? The importance of perceived trust and the usefulness of AI-powered retail platforms in shaping purchase intention\u0026rdquo;, \u003cem\u003eYoung Consumers\u003c/em\u003e, Vol. 25 No. 6, pp.969\u0026ndash;989, doi: doi.org/10.1108/YC-02-2024-1991\u003c/li\u003e\n\u003cli\u003eField, H. (2025), \u0026ldquo;Perplexity AI wrapping talks to raise $500 million at $14 billion valuation\u0026rdquo;, available at: https://www.cnbc.com/2025/05/12/perplexity-funding-round-comet.html (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003eGartner (2023), \u0026ldquo;Predicts 2024: how GenAI will reshape tech marketing\u0026rdquo;, available at: https://www.gartner.com/en/doc/800771-predicts-2024-how-genai-will-reshape-tech-marketing (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003eGhaffari, G., Yousefimehr, B. and Ghatee, M. (2024), \u0026ldquo;Generative-AI in e-commerce: use-cases and implementations\u0026rdquo;, in IEEE (Ed.), \u003cem\u003e2024 20th CSI International Symposium on Artificial Intelligence and Signal Processing (AISP 2024), Babol, Iran, 21\u0026ndash;22 February 2024\u003c/em\u003e, IEEE, Piscataway, NJ, pp.1\u0026ndash;5, doi: doi.org/10.1109/AISP61396.2024.10475266\u003c/li\u003e\n\u003cli\u003eGuevara, W. (2025), \u0026ldquo;Google AI Overviews: new CTR study reveals how to navigate negative SERP impact\u0026rdquo;, available at: https://www.amsive.com/insights/seo/google-ai-overviews-new-research-reveals-how-to-navigate-click-drop-off/ (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003eGuha, A., Grewal, D., Kopalle, P.K., Haenlein, M., Schneider, M.J., Jung, H., Moustafa, R., Hegde, D.R. and Hawkins, G. (2021), \u0026ldquo;How artificial intelligence will affect the future of retailing\u0026rdquo;, \u003cem\u003eJournal of Retailing\u003c/em\u003e, Vol. 97 No. 1, pp.28\u0026ndash;41, doi: doi.org/10.1016/j.jretai.2021.01.005\u003c/li\u003e\n\u003cli\u003eGupta, A.S. and Mukherjee, J. (2025), \u0026ldquo;Framework for adoption of generative AI for information search of retail products and services\u0026rdquo;, \u003cem\u003eInternational Journal of Retail \u0026amp; Distribution Management\u003c/em\u003e, Vol. 53 No. 2, pp.165\u0026ndash;181, doi: doi.org/10.1108/IJRDM-05-2024-0203\u003c/li\u003e\n\u003cli\u003eGuo, J. and Lv, Z. (2022), \u0026ldquo;Application of digital twins in multiple fields\u0026rdquo;, \u003cem\u003eMultimedia Tools and Applications\u003c/em\u003e, Vol. 81, pp.26941\u0026ndash;26967, doi: doi.org/10.1007/s11042-022-12536-5\u003c/li\u003e\n\u003cli\u003eHaigler, N. (2024), \u0026ldquo;How AI overviews are impacting CTR: 5 initial takeaways\u0026rdquo;, available at: https://www.seerinteractive.com/insights/how-ai-overviews-are-impacting-ctr-5-initial-takeaways (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003eHuh, J. and Kim, H. (2025), \u0026ldquo;\u0026lsquo;Function or emotion?\u0026rsquo;: understanding smart shopping via voice AI based on the LVSEM approach\u0026rdquo;, \u003cem\u003eInternational Journal of Retail \u0026amp; Distribution Management\u003c/em\u003e, Vol. 53 No. 7, pp.623\u0026ndash;638, doi: doi.org/10.1108/IJRDM-07-2024-0370\u003c/li\u003e\n\u003cli\u003eJan, I. U., Ji, S. And Kim, C. (2023) \u0026quot;What (de) motivates customers to use AI-powered conversational agents for shopping? The extended behavioral reasoning perspective\u0026quot;, \u003cem\u003eJournal of Retailing and Consumer Services, \u003c/em\u003eVol. 75, doi: doi.org/10.1016/j.jretconser.2023.103440\u003c/li\u003e\n\u003cli\u003eJung, Y., Chen, C., Jang, E. and Sundar, S.S. (2024), \u0026ldquo;Do we trust ChatGPT as much as Google Search and Wikipedia?\u0026rdquo;, in Mueller, F.F., Kyburz, P., Williamson, J.R. and Sas, C. (Ed.s), \u003cem\u003eCHI 2024 \u0026mdash; Extended Abstracts of the 2024 CHI Conference on Human Factors in Computing Systems, \u003c/em\u003eAssociation for Computing Machinery, New York, NY, pp.1\u0026ndash;9, doi: doi.org/10.1145/3613905.3650862\u003c/li\u003e\n\u003cli\u003eLaw, R. and Guan, X. (2025), \u0026ldquo;AI overviews reduce clicks by 34.5%\u0026rdquo;, available at: https://ahrefs.com/blog/ai-overviews-reduce-clicks/ (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003eLopes, J.M., Silva, L.F. and Massano-Cardoso, I. (2024), \u0026ldquo;AI meets the shopper: psychosocial factors in ease of use and their effect on e-commerce purchase intention\u0026rdquo;, \u003cem\u003eBehavioral Sciences\u003c/em\u003e, Vol. 14 No. 7, p.616, doi: doi.org/10.3390/bs14070616\u003c/li\u003e\n\u003cli\u003eMacLachlan, R.J.R., Adams, R., Lauro, V., Murray, M., Magueijo, V., Flockhart, G. and Hasty, W. (2024), \u0026ldquo;Chat-GPT: a clever SE or a creative design assistant for students and industry?\u0026rdquo;, in Grierson, H., Bohemia, E. and Buck, L. (Ed.s), \u003cem\u003eDS 131: Proceedings of the International Conference on Engineering \u0026amp; Product Design Education (E\u0026amp;PDE 2024): Rise of the Machines: Design Education in the Generative AI Era, Aston University, Birmingham, \u003c/em\u003eUK, pp.414\u0026ndash;419, doi: doi.org/10.35199/epde.2024.70\u003c/li\u003e\n\u003cli\u003eMari, A., Mandelli, A. and Algesheimer, R. (2024), \u0026ldquo;Empathic voice assistants: enhancing consumer responses in voice commerce\u0026rdquo;, \u003cem\u003eJournal of Business Research\u003c/em\u003e, Vol. 175, p.114566, doi: doi.org/10.1016/j.jbusres.2024.114566\u003c/li\u003e\n\u003cli\u003eMehta, I. (2025), \u0026ldquo;Google introduces Shop with AI Mode with price tracking, agentic checkout, and virtual try-on\u0026rdquo;, available at: https://techcrunch.com/2025/05/20/google-adds-ai-powered-shopping-features-for-discovery-and-easy-check-out/ (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003eMehta, R. and Chilimbi, T. (2024), \u0026ldquo;Amazon announces Rufus, a new generative AI-powered conversational shopping assistant\u0026rdquo;, available at: https://www.aboutamazon.com/news/retail/amazon-rufus (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003eMpinganjira, M., Dlodlo, N. and Idemudia, E.C. (2024), \u0026ldquo;Perceived experiential value and continued use intention of e-retail chatbots\u0026rdquo;, \u003cem\u003eInternational Journal of Retail \u0026amp; Distribution Management\u003c/em\u003e, Vol. 52 No. 13, pp.121\u0026ndash;135, doi: doi.org/10.1108/IJRDM-04-2023-0237\u003c/li\u003e\n\u003cli\u003eOpenAI (2025), \u0026ldquo;Help ChatGPT discover your products\u0026rdquo;, available at: https://openai.com/chatgpt/search-product-discovery/ (accessed 29 August 2025).\u003c/li\u003e\n\u003cli\u003ePandya, V. (2025), \u0026ldquo;Adobe Analytics: traffic to U.S. retail websites from generative AI sources jumps 1200 percent\u0026rdquo;, available at: https://blog.adobe.com/en/publish/2025/03/17/adobe-analytics-traffic-to-us-retail-websites-from-generative-ai-sources-jumps-1200-percent (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003ePantano, E. and Pizzi, G. (2020), \u0026ldquo;Forecasting artificial intelligence on online customer assistance: evidence from chatbot patents analysis\u0026rdquo;, \u003cem\u003eJournal of Retailing and Consumer Services\u003c/em\u003e, Vol. 55, doi:doi.org/10.1016/j.jretconser.2020.102096\u003c/li\u003e\n\u003cli\u003ePerplexity (2024), \u0026ldquo;Shop like a pro: Perplexity\u0026rsquo;s new AI-powered shopping assistant\u0026rdquo;, available at: https://www.perplexity.ai/hub/blog/shop-like-a-pro (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003ePeruchini, M., Da Silva, G.M. and Teixeira, J.M. (2024), \u0026ldquo;Between artificial intelligence and customer experience: a literature review on the intersection\u0026rdquo;, \u003cem\u003eDiscover Artificial Intelligence\u003c/em\u003e, Vol. 4, p.4, doi: doi.org/10.1007/s44163-024-00105-8\u003c/li\u003e\n\u003cli\u003ePeters, J. (2025), \u0026ldquo;ChatGPT is getting better for shopping\u0026rdquo;, available at: https://www.theverge.com/news/656729/openai-chatgpt-search-shopping (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003eResearch and Markets (2025), \u0026ldquo;SE Optimization Services Market Report 2025\u0026rdquo;, available at: https://www.researchandmarkets.com/report/seo (accessed 18 June 2025).\u003c/li\u003e\n\u003cli\u003eRhydderch, A. (2017), \u0026ldquo;Scenario building: the 2x2 matrix technique\u0026rdquo;, in Bourse, F. and Ro\u0026euml;ls, C. (Ed.s), \u003cem\u003eProspective and Strategic Foresight Toolbox\u003c/em\u003e, Futuribles International, France.\u003c/li\u003e\n\u003cli\u003eRincon, L. (2025), \u0026ldquo;Shop with AI Mode, use AI to buy and try clothes on yourself virtually\u0026rdquo;, available at: https://blog.google/products/shopping/google-shopping-ai-mode-virtual-try-on-update/ (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003eScheibenreif, D. and Raskino, M. (2023), \u003cem\u003eWhen Machines Become Customers\u003c/em\u003e, Gartner Inc., Stamford, CT.\u003c/li\u003e\n\u003cli\u003eSelter, J., Fota, A., Wagner, K. and Schramm-Klein, H., 2023, \u0026ldquo;Aspects driving customers\u0026apos; intention to use automated purchasing processes\u0026rdquo; International Journal of Retail \u0026amp; Distribution Management, Vol. 51, No. 9-10, pp. 1158 -1173, doi: doi.org/10.1108/IJRDM-10-2022-0397\u003c/li\u003e\n\u003cli\u003eSeymour, T., Frantsvog, D. and Kumar, S. (2011), \u0026ldquo;History of SEs\u0026rdquo;, \u003cem\u003eInternational Journal of Management \u0026amp; Information Systems (IJMIS)\u003c/em\u003e, Vol. 15 No. 4, pp.47\u0026ndash;58, doi: doi.org/10.19030/ijmis.v15i4.5799\u003c/li\u003e\n\u003cli\u003eShahzad, M. F., Xu, S., An, X. And Javed, I. (2024), \u0026quot;Assessing the impact of AI-chatbot service quality on user e-brand loyalty through chatbot user trust, experience and electronic word of mouth\u0026quot;, \u003cem\u003eJournal of Retailing and Consumer\u003c/em\u003e Services, Vol. 79, doi: doi.org/10.1016/j.jretconser.2024.103867\u003c/li\u003e\n\u003cli\u003eSilva S.C., De Cicco R., Vlačić B., and Elmashhara M.G. (2023), \u0026quot;Using chatbots in e-retailing \u0026ndash; how to mitigate perceived risk and enhance the flow experience\u0026quot;. International Journal of Retail \u0026amp; Distribution Management, Vol. 51 No. 3 pp. 285\u0026ndash;305, doi: doi.org/10.1108/IJRDM-05-2022-0163\u003c/li\u003e\n\u003cli\u003eSohn, S., Labrecque, L., Siemon, D. and Morana, S. (2025) \u0026ldquo;Artificial intelligence versus human service agents: How their presence shapes consumer information privacy concerns\u0026rdquo;, \u003cem\u003eJournal of Retailing\u003c/em\u003e, Vol. 101, No. 2, pp. 263-278, doi: doi.org/10.1016/j.jretai.2025.03.003\u003c/li\u003e\n\u003cli\u003eSommerfeld, N., McCurry, M. and Harrington, D. (2025), \u0026ldquo;Goodbye clicks, hello AI: zero-click search redefines marketing\u0026rdquo;, available at: https://www.bain.com/insights/goodbye-clicks-hello-ai-zero-click-search-redefines-marketing/ (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003eSong, M., Xing, X., Duan, Y., Cohen, J., and Mou, J. (2022) \u0026quot;Will artificial intelligence replace human customer service? The impact of communication quality and privacy risks on adoption intention\u0026quot;, \u003cem\u003eJournal of Retailing and Consumer Services\u003c/em\u003e, Vol 66, doi: doi.org/10.1016/j.jretconser.2021.102900\u003c/li\u003e\n\u003cli\u003eStatista (2025), \u0026ldquo;Search advertising \u0026ndash; worldwide\u0026rdquo;, available at: https://www.statista.com/outlook/amo/advertising/search-advertising/worldwide (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003eStrzelecki, A. (2024), \u0026ldquo;Is ChatGPT-like technology going to replace commercial SEs?\u0026rdquo;, \u003cem\u003eLibrary Hi Tech News\u003c/em\u003e, Vol. 41 No. 6, pp.18\u0026ndash;21, doi: doi.org/10.1108/lhtn-02-2024-0026\u003c/li\u003e\n\u003cli\u003eWang, B., Zhou, H., Li, Z., Yang, G., Zheng, P., Song, C., Yuan, Y., Wuest, T., Yang, H. and Wang, L. (2024), \u0026ldquo;Human digital twin in the context of Industry 5.0\u0026rdquo;, \u003cem\u003eRobotics and Computer-Integrated Manufacturing\u003c/em\u003e, Vol. 85, p.102626, doi: doi.org/10.1016/j.rcim.2023.102626\u003c/li\u003e\n\u003cli\u003eWiggers, K. (2025), \u0026ldquo;OpenAI is losing money on its pricey ChatGPT Pro plan, CEO Sam Altman says\u0026rdquo;, available at: https://techcrunch.com/2025/01/05/openai-is-losing-money-on-its-pricey-chatgpt-pro-plan-ceo-sam-altman-says/ (accessed 25 June 2025).\u003c/li\u003e\n\u003cli\u003eYang, Q., Sun, L., Wu, X. and Li, Y. (2025), \u0026ldquo;Implications of retailer-owned digital twins services: the trade-offs between customer experience, misfit returns reduction, and investment costs\u0026rdquo;, \u003cem\u003eJournal of Retailing and Consumer Services\u003c/em\u003e, Vol. 88, doi: doi.org/10.1016/j.jretconser.2025.104478\u003c/li\u003e\n\u003cli\u003eYue, Y., Ng, S. and Basha, N.K. (2024), \u0026ldquo;Consumption values, attitudes and continuance intention to adopt ChatGPT-driven e-commerce AI chatbot (LazzieChat)\u0026rdquo;, \u003cem\u003ePakistan Journal of Commerce and Social Sciences\u003c/em\u003e, Vol. 18 No. 2, pp.249\u0026ndash;284.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8958644/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8958644/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAs Generative Artificial Intelligence (GenAI) like ChatGPT introduces retailing features such as product search, recommendation and checkout, there may be significant impact to retail, e-commerce, digital marketing, and search engines (SE). This study aims to foresight impacts of GenAI applications in such areas, especially e-Commerce. Speculative scenarios derived from signals covering the impact of GenAI on e-commerce, SE, SEO/SEM, shopping assistants, customer purchasing intention, human digital twins, and Large Language Models (LLM) monetization. A 2x2 Scenario Planning Matrix (SPM) speculates how LLM companies (LLMCs) may monetize through advertising or direct sales versus users\u0026rsquo; data trust levels. The SPM proposes a probable \"AI Commerce\u0026rdquo; future where GenAI replaces SE and traditional e-commerce for all online purchasing. Other probable futures include GenAI advertising monetization and potential LLM acquisitions by SE companies. A preferable future with enhanced customers data privacy control is proposed. Diegetic Prototypes (DPs) illustrate probable and preferable futures. Speculative scenarios and DPs help retailers and marketplaces understand GenAI's potential to reshape e-commerce, digital marketing, SEs and online consumer behaviour overall.\u003c/p\u003e","manuscriptTitle":"The Future of Shopping: AI Commerce","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-26 08:25:48","doi":"10.21203/rs.3.rs-8958644/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"b7502dab-d280-40c2-bc03-e6cdf4d48c87","owner":[],"postedDate":"February 26th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-26T19:24:46+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-26 08:25:48","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8958644","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8958644","identity":"rs-8958644","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}