There is a new distribution channel which will give marketers the chance to leapfrog their competition.

As well as continuing selling to people, there will be a race to sell to AI agents, competing for selection in an automated decision process.  This creates new challenges in data quality, integration and demand fulfilment.  The companies with the best cleanest data, strongest trust signals and measurable outcomes will rise to the top.  An AI buyer can compare a brand’s reputation, authenticity, environmental credentials and return rate faster than a human shopper can.

Yet at the same time the pitfalls in modern shopping for humans (the ultimate user) have already multiplied and are set to increase as safeguarding becomes more complex.

Things are already tougher for shoppers.

A research study in October 2025 found that as much as 87% of financial advice on TikTok given out by “finfluencers” was potentially misleading and that only 1 in 7 creators used appropriate disclaimers.

Then there are the “skinfluencers”, especially worryingly some very young ones.  Another study of 1500 children aged between 9 and 12 says that nearly half are using multiple skincare products, with half of those saying that they use them to fix “problem skin”.  Experts say that the marketing of skincare online is leading to a new issue that they are calling “cosmeticorexia” which may cross over into body dysmorphic disorder.

Martin Lewis, the most trusted face in UK financial advice, has spoken out about the huge number of deep fake adverts that he now features in online saying: “these ads – be they for crytpocurrency, investments, Quantum AI – are scams and not genuine.  They’re dangerous and you should be on your guard.”

Caveat emptor, or buyer beware is one of the oldest principles of law.  It originated in Roman law and it pre-dates Christianity.  It is first recorded in English in John Fitzherbert’s 1523 manual on animal care: “The Boke of Husbandry” when the author advises that if you’re buying a horse you should check its condition for yourself.  Some things never change.  If you are buying a used car, you will still want to kick the tyres yourself, or at least ask your good petrol head friend to do so for you.

Conveying good reputation is the original point of a brand.  In the earliest age of industrialisation the brand served as a shortcut to quality credentials in a world of snake oil salesmen and food contamination, back in the 19th century when people moved to cities at scale and no longer knew the vendors of goods personally.  In the last 50 years modern law has seen the obligation shift to the seller in many cases.  Manufacturers and retailers now have a responsibility legally to deliver goods that match the promise. To refrain from claims in advertising that do not match the facts.  To list terms and conditions that might impact on the purchase.

Now, the age of Buyer Beware is about to be upon us again, but for quite different reasons.

Originally a buyer needed to look out because inequality of information.  The seller knows all about what they’re selling.  The buyer must hope that the information shared is fair and accurate, and in the bad old days this was much more random than we have become used to.   

AI is set to disrupt the current status quo.  Not because the AI buyer has imperfect information.  If you ask for a recommendation from AI you will be able to access better information than you would from a traditional search.  No, it is because no-one really knows what AI will do or say, including AI itself and its creators.  (For we must always remember that AI does not understand what it is saying, however convincing it sounds.)

Agentic AI shopping was the subject of a superb panel I joined at SXSW London hosted by international law firm Bird & Bird.  It’s the future and will result in yet more disruption and transformation.  A huge new requirement for safeguards, data security and clarity of liability. An even greater need to keep in mind the principle of Caveat Emptor.  AI works on the basis of existing data and information, and with all the inbuilt biases and prejudices that exist in the world today maybe some buyers might need to be more watchful than others.

Some things are certain.  In the new era we will need to be very clear what it is the humans are good for versus artificial intelligence. Integrity and trust will be paramount.  Buyers will need to beware. Procurement in business will need to understand agentic exchanges. Leaders overseeing the next transformation in business and commerce will need skilful human judgement in critical situations.