ChatGPT ads arrive in the UK, seven months after OpenAI said it would only be testing
Sponsored results reached British users on 11 August. They appear on the Free and Go tiers, they are chosen from what you type, and paying more is the only way to switch them off.
Advertising inside ChatGPT reached the UK on 11 August, alongside Mexico, Brazil, Japan and South Korea. Until now the format had run only in the United States.
The expansion is the part worth noting. When OpenAI announced the format in January it was framed as a test in a single market. Seven months later it is a product with a self-serve campaign manager, and the UK is in it.
Who sees them
Ads appear for logged-in adult users on the Free tier and on ChatGPT Go, the roughly $8-a-month plan OpenAI introduced last year. Plus, Pro, Business, Enterprise and Education subscribers do not see them.
That structure is worth stating plainly, because it is the commercial logic of the whole thing: the advertising is the price of the free tier, and the subscription is the opt-out. Nothing about that is unusual — it is how most of the consumer web works — but it is a change to a product a great many people started using when it had no advertising at all.
OpenAI describes the ads as contextual text placements below the answer, labelled “Sponsored” and visually separated from the response. The company says advertisers do not get access to conversations and that it does not sell user data.
What “contextual” is doing in that sentence
The approach OpenAI has published rests on a distinction that is easy to skate past. Advertisers are not handed your conversations. But the ad shown to you is selected using what you typed.
Search advertising has always worked this way and few people object, because a search query is a short, deliberate statement of intent. A ChatGPT conversation is often neither short nor deliberate. People use it to work through medical worries, money problems, job applications and relationships, in a register much closer to a diary than to a search box.
Targeting on that basis is not the same thing as selling data, and OpenAI is accurate when it says it is not doing the latter. It is a meaningfully different exposure from a search ad, though, and the announcement does not dwell on it.
The UK angle
OpenAI’s earlier disclosures said the format is confined to the Free and Go tiers and excludes under-18s. What has not been set out in detail is how the system distinguishes commercially valuable subjects from sensitive ones, or what happens when a conversation moves from one to the other mid-thread.
British users are covered by UK GDPR and the ICO’s direct marketing rules, which set a higher bar for special-category information — health, and beliefs among it — than US law does. Whether contextual selection inside a conversation engages those rules is genuinely unsettled, and it will be the interesting question of the next year rather than of this week.
Health is where that question bites hardest, and it is not hypothetical: people already bring symptoms to chatbots, and the same companies are building AI that consults rather than merely answers.
We have asked OpenAI whether any subject categories are excluded from ad targeting in the UK, and whether the exclusions differ from the US. We will update this piece if the company responds.