AI assistants have emerged as the key public-facing facet of the most dominant tech movement in recent memory. General-purpose tools like ChatGPT and Google Gemini may be the best known and widely used, but niche-specific iterations and adaptations are arguably more important and impactful from the perspective of individual industries and specific businesses alike.
The launch and evolution of AI shopping assistants serve as the ideal example of what’s happening more broadly with artificially intelligent chat tools at the moment. The obvious advantage of being able to field common customer queries automatically, rather than dedicating human employees to this repetitive task, catalyzed their initial adoption. Now, they can do much more than provide product info and details of return policies.
Distillery shares a rundown of the data underpinning AI shopping assistants so far, illustrating what they offer retail brands, where they might fall short of expectations, and where they’re headed as further changes and improvements arrive.
Appreciating AI’s E-Commerce Influence
The power of AI in an e-commerce context is easy to establish and quantify. Adobe reported that traffic driven to retail sites by AI sources, whether via chat tools or full-blown browsers, was 4,700% higher in the first six months of 2025 than in the same period in the previous year.
One of the driving factors behind this trend, identified by analysts, was consumer trust. 90% of people surveyed said they had faith in the accuracy of results provided by AI tools.
Better yet, from a retailer perspective, is the promise of a 27% decline in bounce rate and a 10% uptick in engagement when visitors click through from an AI source, rather than a traditional search result. The explanation offered is that AI allows prospective customers to arrive better informed and thus in a stronger position to know what to expect from a product landing page, which in turn gives them more incentive to stick around and check that the details they encounter pre-click are accurate.
Exploring the Agentic Angle
Aside from search engine-like research and subsequent on-site purchases, another aspect of AI shopping assistants, which is perhaps even more impactful, is agentic commerce.
A comprehensive overview from McKinsey estimates that as much as $1 trillion in B2C retail revenue could be generated from agentic commerce by the end of the decade. The idea is that AI tools will eliminate friction from the shopping process by handling everything from anticipating what consumers might want to buy to actively placing orders on platforms where these products are available.
Once tools like this become truly site- and platform-agnostic and proven capable of completing comparatively complex transactions, such as those involved in a typical e-commerce purchase, it seems very likely that more and more decisions and touchpoints will be taken out of human shoppers’ hands. The potential for missteps and the erosion of trust that will inevitably result from them remains high, at least until this strand of tech matures. But the potential is huge.
Investigating the Rise of Integration
While the agentic route might be preferable in the long run, the short-term trend is one of direct integration. This provides businesses a way to combine e-commerce with AI shopping assistants directly, effectively cutting out the middle man, which in this case is the increasingly archaic-looking use of a retail website.
Most recently, U.K.-based sportswear brand JD announced a partnership with Microsoft to bring transactional power to the Copilot chatbot for customers in the US. This ability will eventually be extended to Google’s Gemini and ChatGPT. Payment platform Stripe is set to handle the financial side of this, and the announcement gave JD’s shares a 1.2% lift, following a year defined by a declining valuation.
Google has also made headlines following announcements of integrations with Walmart for online shopping purposes. Amazon is taking similar steps, and the prevalence of cross-industry partnerships like this looks set to continue.
Considering the Custom Software Opportunities
While agentic commerce and AI shopping assistants are being merged with broad functionality chatbots, which might be making a splash at the moment, there’s also a key role for on-site tools to play in the development of this trend.
The rise of nearshore software development means more organizations across the retail spectrum, rather than only in the upper echelons in terms of revenues and reach, can justify investing in bespoke tools. This raises the question of how AI-augmented experiences will be created and implemented in e-commerce in the next few years.
What could define these tools, in contrast to their platform-agnostic counterparts, is the richness of features available. For instance, if a fashion retailer can include AI-enabled virtual fitting room features, it will serve the dual purpose of convincing customers to spend more time on its site and simultaneously reducing returns. With goods worth $850 billion returned to merchants last year, this latter benefit could further catalyze the adoption of AI shopping assistants.
The Necessity of Guardrails
The final talking point to touch on regarding AI shopping assistants is one that also applies to AI tools of every type; without guardrails in place, the unpredictability of their behavior threatens to decimate consumer trust in them. It’s easy to imagine a context in which an agentic commerce solution recommends entirely inappropriate products to a user, overorders to the detriment of their bank balance, or even gets manipulated by malicious third parties, causing all sorts of chaos. This is bad for the reputation of AI, but worse for individual retailers, the majority of which are far less able to weather reputational storms than tech giants.
Payment platform providers are leading this charge. Visa’s announcement of its new Trusted Agent Protocol in October last year aims to detect and block malicious bots, recognize and facilitate genuine agentic transactions, and prioritize transparency throughout so that consumers can access all relevant data for each payment made.
Where Mainstream AI Shopping Assistants are Headed
The main takeaway is that AI will continue its march to dominance in e-commerce, and that this journey will have multiple routes and outcomes. From top-level agentic commerce to individual on-site assistant tools, the industry has not seen a shakeup this significant since the start of the online shopping revolution more than three decades ago.
Consumers have already demonstrated their love for the convenience that AI shopping tools offer, and their trust has grown in proportion to their increasing use. For this trend to be sustained, retailers and AI developers must work in unison. Any new tech runs the risk of falling flat if some major negative event occurs, and there’s too much at stake for slip-ups, given the amount invested at the moment.