As our beloved high streets grapple with rising business costs, decaying consumer confidence and yet another bout of tax uncertainty, AI agents are reshaping how consumers shop.
A report published by the Information Commissioner’s Office (ICO) warns that that so-called ‘agentic AI’, a quasi-autonomous form of artificial intelligence, could shift from novelty to norm within five years.
Digital assistants are set to search for deals, manage budgets and go as far as negotiating prices for its user, as well as eventually completing the transaction.
For retailers already struggling with global online giants, the implications are profound.
Unlike today’s chatbots, the likes of ChatGPT or Gemini, agentic AI systems were designed to act with little human oversight.
Rather than waiting to be prompted or questions, they learn preferences and transact, supposedly, in real time.
In retail terms, that means an AI shopper could scan post-Christmas January sales, check a customer’s bank balance, and compare prices across platforms.
It could then place an order based on that information, before the shopper has even had to open an app.
The ICO has dubbed this as the dawn of ‘agentic commerce’, where retailers no longer sell to humans, but to algorithms taking action on their behalf.
For alarmingly then, it is the sun setting on the great British high street that we should be concerned about. To the sector, this presents both a threat and a last roll of the dice.
Retailers must sell to bots
Recent Capgemini research saw that this shift is already underway, reporting that a quarter of customers used generative AI shopping tools in 2025.
What’s more, an additional 31 per cent are planning to do so.
OpenAI, the owner of ChatGPT, now allows its users to buy from retailers like Walmart, or Etsy, directly through the LLM.
Meanwhile, retail giant Amazon has blocked rival AI crawlers, in aim to maintain control of its lucrative ad ecosystem.
“Brands must move beyond being optimised for search, to being optimised for selection,” said Dreen Yang, Capgemini’s global retain head.
That shit requires machine-readable product data to become a main focus, pricing to be constantly updated, and rich contextual detail to be listed on the item.
Take reviews and ratings, long seen as marketing fluff. Those inputs may become decisive into algorithmic decision making.
For bricks and mortar businesses already feeling the squeeze, the transition could be brutal.
But if we lean on the side of optimism, it may offer a form of lifeline where London’s high street retailers, which still enjoy lower vacancy rates than the national average, are given a way to adapt and thrive.
And in a world where 66 per cent of consumers still look for human support when purchasing, especially within luxury retail, the high street’s strengths may lie in combining that physical experience with the rise of this technology.
Privacy fears loom
But the ICO has pointed to the risks that this future may come with.
Such agentic systems may need to access bank accounts or spending histories, as well as other behavioural data to function at its best.
That concentration of personal information raises real questions about consent, transparency and control.
William Malcolm, ICO’s executive director for risk and innovation, said: “While the potential benefits could be transformational, technological advancements must not come at the cost of data privacy”.
The regulator warned that badly designed, open-source systems could overreach those boundaries, processing and consuming more data than it needs.
It could also make opaque decisions that consumers may struggle to challenge.
Throughout the next year, the ICO has warned it will be closely monitoring developments within this space, ensuring privacy guardrails are embedded by design.