The Chatbot Just Walked Into the Mall
Something significant happened in the first week of March 2026. Meta quietly began testing a shopping research feature inside its Meta AI chatbot, making it the latest major tech platform to embed product discovery directly into a conversational AI interface. The move, first reported by Bloomberg on March 3, 2026, puts Meta alongside OpenAI, Google, Perplexity, Amazon, and Microsoft — all of whom have, in the past six months alone, pushed AI-powered shopping tools to their users.
This is not a small announcement buried in a product update changelog. It is a signal that the next battleground in e-commerce is not a search results page, a social feed, or a product listing site. It is a conversation.
For consumers, this shift means product discovery could increasingly happen through a chat interface rather than a browser tab. For brands, it raises urgent new questions about visibility, positioning, and relevance in an AI-mediated buying environment. For advertisers, it opens a debate that hasn’t been fully resolved — one about transparency, bias, and what it actually means when an AI recommends a product.
This post unpacks everything currently known about the Meta AI shopping feature: what it does, how it compares to competitors, what the data says about where AI-driven commerce is actually headed, and what marketers and business owners should be doing about it right now.
What Is the Meta AI Shopping Feature?
According to Bloomberg’s reporting, Meta is testing a shopping research feature within Meta AI — the company’s standalone AI assistant, which is accessible at meta.ai and integrated across Facebook, Instagram, WhatsApp, and Messenger.
The feature, which is currently in a limited rollout to select US-based users of the Meta AI web browser, works like this: a user types a shopping-related query into the Meta AI chat interface. Instead of receiving a text-only response, the user sees a horizontal, scrollable carousel of product images. Each item in the carousel includes:
- The product name and image
- The brand associated with it
- The price
- The website where the product can be purchased
- A brief, bullet-point explanation of why Meta AI is recommending that specific item
There is no in-app checkout. Users who want to buy an item are directed to the merchant’s external website to complete the purchase.
The feature can be personalized. Bloomberg’s testing confirmed that when the chatbot had access to user profile data — specifically location and inferred gender — it tailored responses accordingly. When a Bloomberg reporter tested the tool, asking for puffer jacket recommendations, the chatbot produced results suited to women shopping in New York, based on the profile data Meta already held.
Users who have access to the feature will see a “Shopping research” button inside the query text box when they visit Meta AI on the web. Meta has confirmed it is testing the feature, but has not announced a timeline for a broader rollout.
Context: Why Meta Is Moving Into AI Shopping Now
To understand why Meta is testing this feature, it helps to understand the financial and strategic pressure the company is under.
Meta has committed to spending $600 billion on US-based AI infrastructure over the next three years. With that level of capital expenditure, the company needs to demonstrate that its AI investments translate into revenue — not just user engagement metrics. Meta earns approximately 99% of its revenue from advertising, and AI is already deeply embedded in how it optimizes those ad systems. But advertising on a chatbot interface is a different proposition from advertising in a social feed.
The AI shopping tool represents one of the more straightforward paths to monetization inside a chat environment: surface products, drive clicks to merchant websites, and potentially — though Meta has not confirmed this — collect referral commissions or offer premium placement to advertisers. A Meta spokesperson declined to say whether the company receives referral commissions for the chatbot’s product recommendations and would not comment on whether brands that advertise on Facebook or Instagram receive any priority in the results.
That ambiguity matters, and we will return to it.
Mark Zuckerberg signaled this direction publicly in January 2026, during Meta’s most recent earnings call. He told investors that Meta would soon release products demonstrating its ability to deliver “a uniquely personal experience based on users’ history, interests and relationships.” He also said that Meta’s new agentic shopping tools would “allow people to find just the right, very specific set of products from the businesses in our catalog.” Those comments now look like a preview of exactly what Bloomberg has since confirmed.
Meta AI itself is already at significant scale. The platform crossed 1 billion monthly active users in May 2025 — a milestone reached approximately a year after launch — and is built on Meta’s Llama 4 model, which launched in April 2025. Llama and its derivatives have been downloaded more than 650 million times. The distribution advantage Meta holds over any standalone AI platform is enormous. Every time someone opens Instagram, WhatsApp, or Facebook, Meta AI is potentially within reach.
How Meta’s Shopping Feature Compares to Its Competitors
Meta is entering a space that its rivals have been developing since late 2024 and accelerating through 2025. Here is where each major player currently stands.
OpenAI / ChatGPT
OpenAI launched its Shopping Research feature in November 2025, shortly before the holiday shopping season. The feature operates similarly to Meta’s: users ask for product recommendations, and ChatGPT returns a scrollable carousel of products with prices, links, and contextual notes. OpenAI also launched an Instant Checkout feature in late September 2025, which goes a significant step further — it allows users to complete purchases directly within the ChatGPT interface, without visiting a merchant’s external site. ChatGPT has also integrated with retailers including Walmart, and has a payment partnership with PayPal in development.
Google / Gemini
Google launched its agentic checkout feature for Gemini ahead of the 2025 holiday shopping season. The capability is more advanced than Meta’s current test: Google’s Universal Checkout Platform allows users to log into merchant accounts or pay directly via Google Pay from within the Gemini interface. In January 2026, Google partnered with Walmart to enable conversational shopping through Gemini, allowing customers to browse and purchase Walmart and Sam’s Club products without leaving the chat. Google has also built out an AI Mode in Google Shopping and operates an AI agent that can reportedly call nearby stores to check inventory.
Perplexity
Perplexity relaunched its agentic shopping tool in fall 2025, expanding access beyond its Pro subscribers. The tool builds on Perplexity’s “Buy With Pro” feature and uses memory from past searches to detect shopping intent and personalize results.
Amazon
Amazon’s Rufus, the company’s AI shopping assistant embedded directly in the Amazon app, continues to expand. Amazon also has an “Interests” feature that proactively surfaces products based on stated user preferences. Given that Amazon is both the AI provider and the retailer, its integration is inherently tighter than what Meta or OpenAI can currently offer.
Microsoft
Microsoft is testing Copilot Checkout, which allows users to shop and pay without leaving the chat window — positioning it alongside Google in terms of in-chat transaction capability.
Where Meta Currently Stands
Compared to this field, Meta’s current test is more limited. There is no in-app checkout, no confirmed merchant partnerships announced, and no transparency about how recommendations are ranked or whether advertisers receive preferential placement. Forrester VP and Principal Analyst Sucharita Kodali told CNET that Meta’s move looks like “a copycat move” and an attempt to capitalize on AI shopping momentum, questioning whether it brings anything materially new to merchants or consumers beyond what Instagram already provides.
That is a fair critique at this stage. However, Meta’s potential advantage is its data. The behavioral and social graph data Meta holds on over three billion daily active users across its apps is unlike anything OpenAI or Google has at the same depth within a single platform ecosystem. If Meta can responsibly translate that data into genuinely useful, personalized product recommendations, it could differentiate itself in a crowded market.
The Broader Shift: How Consumers Are Actually Using AI for Shopping
The excitement around AI shopping tools in tech circles does not yet fully match consumer behavior on the ground — but the trajectory is clear.
A September 2025 survey by CivicScience and EMARKETER found that just 2% of US digital buyers started their online product searches with an AI assistant or chatbot. That number is low. But the same research infrastructure that produced that figure also tracks behavioral shifts quickly, and the data from early 2026 is notably different.
As of January 2026, 41% of consumers had used dedicated AI platforms for product discovery, according to PYMNTS Intelligence. More notably, 33% said they had fully replaced their prior methods — not layered AI on top of Google search, but abandoned the old approach entirely. Among Gen Z and power users, more than a third now start their daily tasks on AI platforms first.
BCG’s research shows that shopping-related generative AI use grew by 35% between February and November 2025. A September 2025 survey by PayPal found that 61% of Gen Z shoppers used AI tools to help with a purchase in the last year.
The 2025 holiday season provided a meaningful data point at scale. AI agents influenced 20% of global retail sales during that period, driving an estimated $262 billion in purchases through high-intent product discovery — a figure from Envive AI’s research on AI chatbot and agent statistics.
The conversion dynamics are also compelling. Data from Neuwark’s conversational commerce analysis shows that shoppers who interact with an AI chatbot during a session convert at a rate of 12.3%, compared to 3.1% for those who do not — nearly a four-times lift.
eMarketer projects that AI platforms will account for 1.5% of total retail e-commerce sales in 2026, or approximately $20.9 billion in spending. That same firm projects AI platform-driven e-commerce sales will surpass $144 billion by 2029, representing 8.8% of total retail e-commerce.
The AI shopping assistant market specifically — the segment that encompasses tools like Meta’s new feature — was valued at $3.36 billion in 2024 and is projected to grow to $28.54 billion by 2033, a compound annual growth rate of nearly 27%, according to Grand View Research. North America currently holds 40% of this global market.
The conversational commerce market, which is broader and includes all purchase-related interactions through messaging and voice interfaces, sits at $12.64 billion in 2026 and is expected to reach $22.56 billion by 2031, according to Mordor Intelligence. Other analysts peg the number differently — Future Market Insights values it at $10.1 billion in 2026 with a projected rise to $40.2 billion by 2036.
The numbers vary by methodology, but every estimate points in the same direction.
The Trust and Bias Problem Every AI Shopping Platform Faces
The commercial promise of AI shopping tools is real. The trust problem is equally real.
A Gartner survey cited by Brad Jashinsky, Director Analyst at Gartner, found that nearly two-thirds of consumers believe generative AI-powered shopping tools — whether provided by retailers or by AI platforms — will make biased shopping recommendations. That is a majority of consumers entering an AI-assisted shopping interaction already skeptical that the recommendations they receive are neutral.
A YouGov survey published in 2025 found that just 26% of Americans trust AI in retail. Among those who do engage, the level of trust granted is cautious: few are willing to let AI place orders autonomously or handle service issues without human oversight.
The trust deficit is especially acute for a platform like Meta, which carries significant accumulated skepticism around data privacy and the integrity of algorithmic recommendations from its years operating Facebook and Instagram. The fact that Meta has declined to clarify whether brands that already advertise on its platforms receive preferential placement in the shopping chatbot results is not a minor disclosure gap. It is the central question consumers and marketers will want answered before treating Meta AI’s shopping results as genuinely objective.
This is not a problem unique to Meta. Gartner’s Jashinsky noted that more than half of customer experience leaders fear AI delivering inaccurate responses or content that damages the customer experience. And he pointed out that consumer behavior changes slowly. Mobile e-commerce didn’t surpass desktop e-commerce in the US until 2023 — 16 years after the iPhone launched. The behavioral shift to AI-first shopping will likely be measured in years, not quarters.
For Meta’s feature to earn trust at scale, the company will need to address several open questions: How are products selected for the carousel? Does advertising spend on Facebook or Instagram influence placement? Are the bullet-point explanations for recommendations generated independently of commercial relationships? The opacity around these questions is a strategic liability.
What This Means for Brands and Retailers
Whether or not Meta’s current test evolves into a full product release, the direction is clear: AI interfaces are becoming a product discovery channel that brands cannot afford to ignore.
The implications for brand visibility are significant and still being worked out. Traditional SEO has been the discipline that ensures brands appear prominently in text-based search results. The emerging discipline — variously called GEO (Generative Engine Optimization), AEO (Answer Engine Optimization), or AI SEO — is concerned with ensuring that brands appear prominently in AI-generated responses and product carousels.
A February 2026 article in the New York Times captured this dynamic with the headline “Chatbots Are the New Influencers Brands Must Woo” — pointing to a growing realization among marketing teams that they need to win over AI recommendation systems the same way they’ve historically tried to win over human reviewers, influencers, and search algorithms.
For brands and retailers, the practical implications of the Meta AI shopping feature and its equivalents across ChatGPT, Gemini, and Perplexity include:
Product Data Quality: AI shopping tools pull structured product data including images, prices, and descriptions. Brands with incomplete, inconsistent, or outdated product data across the web will struggle to surface in AI-generated carousels, regardless of how strong their paid advertising or SEO presence is. Clean, accurate, and well-structured product feeds are now a baseline requirement.
Review Ecosystems and Authority Signals: AI tools weigh third-party reviews, editorial mentions, and authority signals when generating recommendations. Brands with strong review profiles across credible platforms are more likely to appear in AI-curated results. This means that reputation management is now directly linked to AI discoverability.
Conversational Discoverability: The queries users type into Meta AI for shopping research are phrased very differently from the keyword searches they type into Google. Users might ask “what’s a good lightweight hiking boot for wide feet under $150” rather than “lightweight hiking boots men.” Brands that understand how their products are described in natural language — and ensure that language appears in their product content, reviews, and editorial coverage — will be better positioned in AI results.
Merchant Relationships: Google and OpenAI have already established formal partnerships with retailers like Walmart, Shopify, and Target. Meta has not announced equivalent partnerships. As this space matures, the depth and formality of merchant-AI platform relationships will likely shape which products and brands appear most frequently.
Checkout Friction: Meta’s current test requires a click-through to external merchant sites. ChatGPT and Google are further along in reducing that friction with in-chat checkout. For brands, minimizing the steps between an AI recommendation and a completed purchase will become a competitive differentiator.
Revenue Models, Advertising, and the Question of Paid Placement
The economic logic of AI shopping features is straightforward in theory and complicated in practice.
The two most discussed monetization paths are sponsored placements — where brands pay to appear in AI-generated shopping carousels, similar to sponsored listings in a Google Shopping result — and referral commissions, where the AI platform earns a percentage of each transaction it facilitates.
OpenAI has not been transparent about the commercial relationships underpinning ChatGPT’s shopping results. Google has established formal retailer partnerships, and questions about those relationships have reached the political level — Senator Elizabeth Warren wrote to Google in early 2026 asking for clarity about whether sponsored products are disclosed when they appear through Gemini’s shopping integrations.
Meta, as noted, has declined to comment on either referral commissions or advertiser prioritization. This silence is unlikely to be sustainable as the feature scales. The company that earns nearly all of its revenue from advertising has an obvious commercial interest in connecting its advertising products to its AI recommendations. Whether it can do so in a way that maintains user trust — or whether, as the Social Media Today analysis suggests, “the responses could be viewed as being skewed towards paying ad partners” — is the central tension Meta must navigate.
The parallel with Google is instructive. Google’s search advertising model was built on the assumption that paid placements (labeled as ads) and organic results (not paid) could coexist without destroying user trust in the organic results. That tension has played out for 25 years. Meta is now attempting something similar in a conversational AI context, where the user experience is fundamentally different and where the labeling of paid versus organic content is less established.
Carousel formats, like those Meta is using, offer one potential mitigation: by presenting multiple products rather than a single recommendation, they give users choice and reduce the visibility of any single sponsored placement. This is likely a deliberate design decision.
Meta’s Infrastructure Bet and What It Means for AI Product Scale
Any analysis of Meta’s AI shopping ambitions needs to be grounded in the scale of investment behind them. The $600 billion infrastructure commitment is not a figure to read past. At that level of spending over three years, Meta is building AI infrastructure at a scale that dwarfs most national technology programs. The computing resources being deployed are intended to support AI capabilities at the scale of Meta’s three-billion-plus daily user base.
This investment context matters for the shopping feature because the quality of AI-powered product recommendations is directly tied to computational capacity — the ability to process user queries, cross-reference product databases, apply personalization filters, and generate coherent explanations, all in real time, across hundreds of millions of simultaneous interactions.
Meta’s Llama 4 model, which powers Meta AI, is an open-weight model available to developers and deployed at scale within Meta’s own products. Llama 4 Scout, one of the models in the Llama 4 family, offers a 10-million-token context window — larger than most closed-source alternatives. That capacity supports the kind of deep personalization Zuckerberg has described: a model that knows a user’s preferences, history, and social context well enough to make genuinely useful, individualized recommendations rather than generic carousel results.
What Marketers and Business Owners Should Do Right Now
The Meta AI shopping feature is in testing, not full deployment. But the trend it represents is already underway across multiple platforms. Here is a practical framework for brands and marketers:
1. Audit your product data infrastructure. AI shopping carousels rely on structured data: accurate titles, current prices, high-quality images, clear product descriptions, and verified merchant information. Run an audit of how your products appear in third-party databases, merchant feeds, and review platforms. Gaps here will translate into gaps in AI visibility.
2. Build out your review and authority profile. AI tools that generate product recommendations do not rely solely on advertiser data. They draw on public signals: review volumes, rating averages, editorial mentions, and domain authority. A brand with strong, legitimate third-party endorsements across credible channels is better positioned in AI-generated results than one that relies primarily on paid placements.
3. Map your products to natural language queries. Think about how customers describe their problems, needs, or purchase intentions in conversational terms. “What should I buy for a first-time surfer?” or “Which laptop is best for college students who do a lot of video editing?” These are the types of queries entering AI shopping interfaces. Your product content — descriptions, FAQs, editorial copy — should reflect this language.
4. Monitor where your products appear in AI results. ChatGPT, Gemini, Perplexity, and now Meta AI are all live or testing shopping features. Systematically query these tools as a consumer would, and track whether and how your products appear. This is nascent territory, but brands that establish a monitoring practice now will have a head start on optimization.
5. Develop a position on AI shopping advertising. As these platforms develop sponsored placement models, your brand needs a considered stance. What is your budget framework for AI platform advertising? How do you evaluate performance when attribution is complex and the purchase completes on an external site? These questions require planning before the commercial products are finalized.
6. Watch Meta’s advertiser communications closely. Meta’s transition from confirming a test to rolling out a commercialized feature will include communications to its advertiser base. If you are an active Meta advertiser, understanding how shopping placements integrate with your existing campaigns will be important for budget planning.
The Road Ahead for AI-Mediated Commerce
The launch of Meta’s AI shopping feature test, taken alongside the ChatGPT, Gemini, and Perplexity developments of the past six months, confirms that AI-mediated product discovery is moving from experiment to infrastructure. The category is not yet dominant. The 2% of US digital buyers who currently start product searches on AI platforms are the early adopters, not the mainstream. But the 41% who have tried AI for product discovery as of early 2026, and the 33% who have replaced their old methods, represent a wave that is accelerating.
Consumer behavior will lag behind technological capability, as it always does. Brad Jashinsky’s point about mobile e-commerce taking 16 years to surpass desktop is worth keeping in mind. The pace of change in the AI era may be faster, but the fundamental dynamic — people adopting new tools when they demonstrably improve their experience — applies here too.
The platforms that build AI shopping tools people actually trust, that surface products people actually want, and that provide enough transparency to sustain credibility over time will shape this next chapter of commerce. The ones that optimize narrowly for short-term monetization without addressing the bias and transparency questions will see the trust erosion that erodes long-term engagement.
Meta enters this race with the largest captive user base of any platform testing these features, and with data assets its competitors cannot easily replicate. What it does not yet have is a clear, public answer to the transparency question — and in a space where two-thirds of consumers already expect biased recommendations, clarity on that point will be a competitive asset, not just a regulatory obligation.
The feature is in testing. The stakes are not.
Frequently Asked Questions (FAQ): Meta AI Shopping Tool
What exactly is the Meta AI shopping feature? The Meta AI shopping feature is an experimental function being tested within Meta’s AI chatbot platform, Meta AI. When users ask the chatbot for product suggestions, it responds with a scrollable carousel of products that includes the product image, brand name, price, the website where the item is available for purchase, and a short explanation of why each product was recommended. The feature is currently being tested with select US-based users who access Meta AI through a web browser at meta.ai.
Is the Meta AI shopping feature available to all users? No. As of March 2026, the feature is in a limited test phase available to select US users on the Meta AI web browser. Meta has confirmed the test but has not provided a timeline for a broader rollout. Users with access will see a “Shopping research” button inside the text input box on the Meta AI web interface.
Can I buy products directly within Meta AI? No. Meta AI’s shopping feature does not currently support in-app checkout or direct purchases. When a user wants to buy a product surfaced by the chatbot, they must click a link to the merchant’s external website to complete the transaction. This distinguishes Meta’s current feature from Google Gemini’s Universal Checkout and OpenAI’s Instant Checkout, both of which offer more advanced in-chat purchasing capabilities.
How does Meta AI decide which products to recommend? Meta has not publicly disclosed the full methodology behind its product recommendations. The available information from Bloomberg’s testing confirms that the chatbot uses publicly available product data including brand, price, and merchant information, and applies personalization based on data Meta already holds about the user — including location and inferred demographic information. Meta has declined to comment on whether brands that advertise on Facebook or Instagram receive preferential placement in shopping results, and has not confirmed whether it earns referral commissions.
How does Meta AI’s shopping feature compare to ChatGPT’s? Both show scrollable carousels of product recommendations with prices and merchant links. OpenAI’s ChatGPT Shopping Research launched in November 2025 and operates similarly, but ChatGPT has gone further with Instant Checkout (launched September 2025), which allows users to complete purchases directly within the chat interface. ChatGPT also has established retailer integrations with Walmart and a PayPal payment partnership in development. Meta’s current feature does not yet match these capabilities.
How does Meta AI shopping compare to Google Gemini? Google Gemini’s AI shopping capabilities are currently more advanced than Meta’s test feature. Gemini offers agentic checkout via Google Pay, allowing purchases to be completed without leaving the chat interface. Google has partnered with Walmart, Shopify, and Wayfair to enable direct browsing and purchasing within Gemini. Google’s AI Mode in Google Shopping and its agent that can call stores to check inventory represent a more fully developed commercial product. Meta’s test is at an earlier stage.
What data does Meta use to personalize shopping recommendations? According to Bloomberg’s reporting, Meta AI personalizes shopping results using data it already holds about users, including location and the gender it infers from the user’s name or profile information. Meta did not disclose the full scope of what data it draws on for personalization, but given the depth of behavioral, social, and preference data Meta accumulates across Facebook, Instagram, WhatsApp, and Messenger, the potential for personalization is substantial.
Will Meta AI shopping show ads? Meta has not confirmed a commercial model for the shopping feature at this stage of testing. The company has not said whether it plans to include sponsored placements in shopping carousels or whether existing Meta advertisers will have access to prioritized placement. Given that advertising represents approximately 99% of Meta’s revenue, it is widely expected that a monetization model will be developed, but no details have been officially disclosed.
Is there a privacy concern with Meta using my data for shopping recommendations? Meta’s use of existing user profile data for shopping personalization raises relevant privacy questions. The company’s data practices are governed by its privacy policy and applicable regulations including GDPR in Europe and CCPA in California. Users who are uncomfortable with personalized recommendations based on their profile data have the option to adjust their privacy settings within Meta’s apps, though the specific controls available for the AI shopping feature have not been publicly detailed during the test phase.
How many people use Meta AI? Meta AI crossed 1 billion monthly active users in May 2025. It is available across Meta’s family of apps — Facebook, Instagram, WhatsApp, and Messenger — as well as through a dedicated web interface at meta.ai and a standalone app. Meta AI is available in more than 60 countries and supported in multiple languages.
What is the market opportunity for AI shopping assistants? The AI shopping assistant market was valued at approximately $3.36 billion in 2024 and is projected to grow to $28.54 billion by 2033, representing a compound annual growth rate of nearly 27%, according to Grand View Research. North America holds approximately 40% of the global market share. The broader conversational commerce market — which includes all commerce conducted through messaging and AI interfaces — is valued at $12.64 billion in 2026 and projected to grow to $22.56 billion by 2031 (Mordor Intelligence). eMarketer projects AI platform-driven retail e-commerce sales will exceed $144 billion by 2029.
Are consumers actually using AI to shop? Adoption is growing but still early-stage. A September 2025 CivicScience/EMARKETER survey found that just 2% of US digital buyers started product searches with an AI assistant — but by January 2026, 41% of consumers had used dedicated AI platforms for product discovery, and 33% said they had fully replaced their previous methods. Among Gen Z, 61% used AI tools to help with a purchase in the last year (PayPal survey, September 2025). During the 2025 holiday season, AI agents are estimated to have influenced 20% of global retail sales.
Do consumers trust AI shopping recommendations? Trust is a known challenge. A Gartner survey found that nearly two-thirds of consumers believe generative AI-powered shopping tools will make biased recommendations. YouGov data from 2025 found that only 26% of Americans trust AI in retail. Consumer trust is higher among younger demographics, particularly Gen Z, but remains a significant barrier to mainstream adoption across all age groups.
What should brands do to appear in Meta AI shopping results? While Meta has not published optimization guidelines for the shopping feature, the general principles of AI visibility apply: maintain accurate and complete product data across merchant feeds and third-party databases, build strong review profiles on credible platforms, ensure your product content reflects natural language search patterns, and monitor how your products appear in AI chatbot queries across Meta AI and competing platforms. As the feature develops and a commercial model is introduced, additional optimization and advertising opportunities will likely become available.
Which other AI platforms have shopping features? Several major platforms now have AI shopping capabilities. OpenAI’s ChatGPT launched Shopping Research in November 2025 and Instant Checkout in September 2025. Google Gemini launched agentic checkout before the 2025 holiday season and has partnerships with Walmart, Shopify, and Wayfair. Perplexity relaunched its agentic shopping tool in fall 2025 with expanded access and PayPal integration. Amazon’s Rufus is embedded in the Amazon app. Microsoft is testing Copilot Checkout. Meta is now the latest to enter the space.
What is agentic shopping and how does it differ from AI product search? Agentic shopping refers to AI systems that can take actions on a user’s behalf — not just recommend products, but actually complete transactions, compare prices across retailers, or take follow-up steps in a purchase journey. Meta’s current test is closer to AI product search: it surfaces recommendations but stops short of facilitating transactions within the interface. Zuckerberg’s public statements have referenced agentic shopping tools as a goal, suggesting the current test is a step toward a more capable feature.
What is Meta’s broader AI strategy? Meta’s AI strategy is built on three interconnected elements: Llama (its open-weight model family, now at version 4), Meta AI (its consumer-facing AI assistant), and AI infrastructure (the $600 billion infrastructure investment over three years). The shopping feature is part of a broader push to make Meta AI the default personal assistant for its three-billion-plus daily active users — capable of tasks spanning productivity, creative work, information retrieval, and increasingly, commercial transactions. Zuckerberg has framed the ambition as building “personal superintelligence” that can deliver uniquely tailored experiences.
When will the Meta AI shopping feature be widely available? Meta has not announced a timeline for a broader rollout. The feature is currently in a limited test with select US users on the Meta AI web browser. Given that ChatGPT and Gemini have already launched their shopping features publicly, competitive pressure suggests Meta will move toward a broader rollout in 2026, though no specific date has been confirmed.
Wrapping It Up
The Meta AI shopping tool test is one data point in a much larger pattern. Every major AI platform is building toward a world where product discovery, research, comparison, and purchase can all happen inside a single conversational interface. Meta’s entry into this space, backed by a billion-user distribution platform and years of behavioral data, represents a meaningful development — not because the current test is technically ahead of its competitors, but because the scale at which Meta operates means that even an early-stage feature has the potential to reshape shopping behavior for a significant share of the global market.
The platforms that define AI-mediated commerce will be those that get three things right simultaneously: accurate and genuinely useful recommendations, a transparent and trustworthy commercial model, and a checkout experience smooth enough to reduce the friction that currently exists between an AI suggestion and a completed transaction. Meta is working on all three. How quickly it closes the gap with more advanced competitors — and how honestly it answers the transparency questions it has so far declined to address — will determine whether this becomes one of the company’s most valuable products or another feature that fails to translate technical capability into consumer trust.
Brands and marketers who treat this as a future concern are already behind. The infrastructure being built today will determine who is visible — and who is invisible — in the AI shopping interfaces that billions of consumers will use over the next decade.
About ALM Corp
ALM Corp is a digital strategy and AI marketing consultancy helping brands navigate the rapidly shifting landscape of AI-driven commerce, search, and consumer engagement. As AI platforms like Meta AI, ChatGPT, and Google Gemini become primary surfaces for product discovery, ALM Corp works with businesses to ensure they are visible, credible, and competitive in these emerging environments.
Our services span AI Search Optimization (ASO/GEO), conversational commerce strategy, product data architecture, AI content strategy, and paid media planning across social and AI platforms — including Meta’s advertising ecosystem. We have published extensively on topics including ChatGPT Shopping Research, AI-mediated product discovery, and the shift from keyword search to conversational discovery.
If your brand is asking how to appear in AI shopping carousels, how to structure your product content for AI visibility, or how Meta’s evolving ad products interact with its AI features, ALM Corp has the expertise to help you build a strategy before the competitive window closes. Contact us to start the conversation.
