Artificial intelligence is becoming increasingly present in the business world, with more shoppers using the tech to navigate purchases. But what does this mean for the future of commerce?
Greater uptake of AI is forcing a number of sectors into a rapid transition, and shopping is no exception.
A quarter of consumers used generative AI shopping tools in 2025 and a further 31% plan to adopt them in future.
That’s according to a new report from IT firm Capgemini, based on global consumer responses and interviews with retail executives.
“Brands must move beyond being optimised for search to being optimised for selection,” said Dreen Yang, EVP, global consumer products and retail lead at Capgemini. “Success increasingly depends on being chosen by algorithms, not just found by consumers.”
In practice, shoppers can apply AI in different ways, whether they make purchases directly through a chatbot, use algorithms for advice, or allow a virtual assistant to spend money on their behalf.
Last year, for instance, OpenAI announced that it would be allowing US users to purchase Etsy, Shopify, and Walmart goods through ChatGPT.
Other companies, such as Amazon, have proven more hesitant, blocking OpenAI crawlers and thereby restricting access to its product listings. By focusing on its own AI ambitions, Amazon hopes to retain control over the e-commerce advertising market.
According to Capgemini, 63% of surveyed consumers want generative AI to provide hyper-personalised shopping content, although concerns around the technology remain. For instance, 71% of respondents said they were worried about how generative AI uses their information.
The desire for human assistance also remains strong, with 66% of those surveyed valuing human support at the point of purchase. This means brands should offer “a balanced mix of digital convenience and human support,” said Capgemini.
In particular, helpline chatbots on company sites, intended to give instant, 24/7 support, were comparatively unpopular, only satisfying 57% of respondents.
Appealing to the bots
According to the report, retailers must adapt to the new age of shopping or risk being left behind.
To improve traction with AI bots, firms should provide machine-readable, continuously refreshed data, enriched with product attributes. This means that the technology can interpret, surface, and recommend products in real-time.
AI tools are more likely to recommend companies that give detailed context about their goods, as it is easier for the bot to understand who the item is for and why it is relevant.
Third-party sources like reviews and ratings also matter, said Capgemini, as they influence how large language models describe the goods or the brand.
Looking ahead, the continued popularity of chatbot shopping is, however, reliant on a number of factors. One key question revolves around the appetite of AI firms for advertisement. If shopping assistants become densely populated with adverts, it’s possible that consumer trust in their recommendations will decline.
Tapping into the transition
When it comes to a firm’s own AI offerings, it's important to adapt the tech to the consumer’s preferences, said Capgemini.
In other words, brands that stand out are offering diverse shopping experiences through one interface. Some users like to explore on their own, while others prefer step-by-step recommendations, both of which can be accommodated using AI tools. Novel offerings include “try-on” features for clothes or accessories, as well as ‘AI style advisers”.
Transparency and control around this technology is also crucial, said the report. A majority of shoppers want to set strict boundaries on digital assistants such as spending caps, and 63% want to approve or reject purchases initiated by AI.
A further 67% want brands to label AI‑generated adverts and content, underlining the desire for human control at every step.
