Editor's note: This brief was summarised by The Property AI Newsroom from a report by The Negotiator. Read the original article for full details.

Buyers risk mistaking AI information for sound judgement, warns buying agent

Homebuyers could cost themselves thousands of pounds by relying on artificial intelligence for their purchasing decisions, according to buying agent Rob Fanshawe, partner at Property Vision. He argues that while AI can provide buyers with one eye "remarkably quickly", "the trouble starts when they convince themselves they have two".

The danger of confident-but-flawed analysis

Fanshawe points out that a buyer considering a house in Paris, London or the Cotswolds can now ask an AI platform almost anything about a property and receive a confident answer in seconds. He illustrates the risks with a hypothetical example: a buyer considering a £6 million house in the Cotswolds asks an AI platform whether the asking price is justified. The AI responds confidently that the property looks expensive against three nearby homes apparently supporting a lower price per square foot.

On closer inspection, however, one of those 'comparables' is an asking price rather than a completed sale, another required a full refurbishment, and the third sits on a busy road. Meanwhile, the genuinely relevant transaction, a similar house sold privately two streets away, never appeared on a portal at all. "The analysis looks impressive. The conclusion is misleading," Fanshawe says.

Why AI research can go wrong

Fanshawe says buyers increasingly arrive "armed with pages of AI-generated research and treat it as a substitute for market knowledge". Sometimes that research is useful, but sometimes it is out of date, based on a flawed comparison, or simply wrong or 'hallucinated'.

Much of the problem, he explains, is that the underlying information AI interprets was imperfect to begin with. AI can confuse asking prices with achieved prices, recycle old commentary as though it describes today's market, or draw sweeping conclusions from postcode-level data that bear little relation to a particular street or house. It can also make weak information sound convincing: buyers today can receive a beautifully structured answer in under 10 seconds and assume uncertainty has disappeared when it has not.

What this means for agents

Fanshawe stresses this is not an argument against buyers being well informed, or against AI. Property Vision uses it themselves. AI can interrogate large quantities of information, summarise documents, research planning histories and flag questions worth investigating further. Used intelligently, it can make an already sophisticated buyer considerably more informed.

But he argues the explosion of readily available information may make experienced advice more valuable, not less. "In the coming years, the advantage in property will not belong to whoever has access to the most information... It will belong to whoever can distinguish intelligence from noise," he says. That means knowing what to trust, what is missing and what actually matters, based on years of market experience. For the foreseeable future, he adds, that remains a very human skill.

For letting agents and inventory clerks, the message is a familiar one: clients arriving with AI-generated summaries may need help separating reliable information from confident-sounding noise.