AI agents are beginning to influence how people discover and buy products, creating a new audience for retailers and consumer brands: the software making recommendations on their behalf. Unilever’s global homecare chief says brands will need to be legible to machines as well as people, without sanding away the character that makes them worth choosing.
Watch Desk analysis
What happened
In a Fortune article, Tati Lindenberg says AI is becoming a new kind of shopping shelf. The piece reports that 52% of UK shoppers have used AI for at least one shopping task in the past three months, citing McKinsey research. Mastercard forecasts that one in 10 people will routinely use AI agents to shop and pay by 2030. Those are different measures: present use of AI for a shopping task is not the same as routinely handing an agent the whole purchase.
There are early signs of the shift. The article says JD Sports lets US shoppers buy through services including Microsoft Copilot, Google Gemini and ChatGPT, while Meta’s shopping-capable Muse agent is expected to launch in the UK “within weeks”. Unilever is working on product information that AI platforms can find and interpret, Lindenberg says.
Why it matters
For an agent to recommend a product, it needs more than a recognisable brand name. Prices, availability, product details and descriptions have to be accessible and clear enough for software to compare. The article reports that nearly half of consumer-products executives surveyed by EY think influencing AI recommendations will be essential to competitiveness over the next five years, while only 20% believe their company can do that today.
That puts a decidedly unglamorous task on the marketing agenda: keep product data accurate, detailed and machine-readable. It also creates a tension. Optimising pages for recommendation systems may help products get noticed, but brands still need the trust and distinctiveness that make people want them in the first place.
Our read
This is a useful glimpse of commerce changing at the information layer, not just through a fresh shopping chatbot. The practical work is product data, pricing and availability that an agent can actually use. But a brand that writes only for machines risks becoming perfectly legible and utterly forgettable. Retailers should make their information easy to parse, then judge the effort by whether agents represent products accurately and shoppers still find the result worth buying.
What to watch
- Whether Meta launches Muse in the UK on the reported timetable.
- Whether retailers publish evidence that AI referrals lead to purchases, not merely visits.
- How brands measure accurate representation across AI shopping services.
Discussion spark: Should brands reshape product information to suit AI shopping agents, or could that make products easier for machines to compare while making the brands themselves less distinctive?
Sources and evidence
- Source update (16 November 2026, 09:00 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.