ChatGPT is testing ads (probably a good move)
Updated 24 September 2026. ChatGPT ads reached Europe on 24 August: 31 countries, Belgium included, on the Free and Go plans. Plus, Pro and Enterprise stay ad-free. A week later the EU designated ChatGPT a Very Large Online Search Engine under the DSA. Corrections are inline.
Watch: Anthropic Claude Super Bowl 2026 ad (YouTube)
Anthropic spent millions on a US audience to make that point. Hours before kickoff, OpenAI proved it by officially rolling out ChatGPT ads.
What ChatGPT ads look like
Since February 9, OpenAI is gradually rolling out labeled "sponsored" links at the bottom of ChatGPT responses:
- Who sees them. Free and Go tier users in the US. Not everyone yet; it's a phased test. (Since 24 August, also in 31 European countries.)
- How they work. Contextual: ask about recipes, you might see a grocery delivery ad.
- Can you dismiss them? Yes. Users can close individual ads or turn off personalization.
Will ChatGPT ads come to Europe?
For sure :)
Update, September 2026: they did, on 24 August.
The ad-carrying ChatGPT Go tier already launched across the EU in January at €8 per month plus VAT. The ads themselves haven't arrived yet. But EU regulation won't be what stops them.
The EU's advertising rules (DSA, DMA, and the AI Act transparency requirements arriving in August 2026) were built for social media and websites, not for an AI chatbot that remembers your conversations and knows not just what you want, but why. No current regulation covers that.
Update, September 2026: now one does, at least in part. On 31 August the Commission made ChatGPT the first chatbot designated a Very Large Online Search Engine under the DSA. OpenAI has four months to comply: yearly independent audits, data access for researchers, and a public repository of the ads it runs. That last one matters for marketers, because your competitors' ChatGPT ads will be public. On consent, OpenAI split it in two. Contextual ads (current chat, city, device, language) are on by default. Personalised ads need an explicit opt-in. Saying no to personalisation doesn't remove the ads.
In any case, when ChatGPT ads will reach Europe, this will be likely the scenario (update: it is exactly how it launched):
- Go (€8/month + VAT): includes ads, no opt-out
- Plus (€23/month + VAT): ad-free
Will people pay to avoid ChatGPT ads? Unlikely.
Every time a tech company has added ads to a subscription product, most users accepted them. OpenAI isn't inventing this pattern. They're following it.
Here are two famous examples:
Netflix launched an ad tier across Europe in November 2022 at €5.99/month in France, similar in Germany, Italy, and Spain. Today, that tier has 190 million monthly active viewers worldwide. Ad tier viewing in Western Europe rose 32% in six months. More than half of new Netflix signups choose the plan with ads.
Uber did the same. You pay for the ride, but since late 2022 you also watch ads. Before, during, and after your trip. Journey Ads now run across ten European markets including France, Germany, Spain, and the Netherlands. Uber's ad business hit $1 billion in annual revenue in 2024. Nobody stopped riding.
And there are many more examples. But Anthropic (Claude) say they will never use ads... Hmm, we'll see.
Can Anthropic win the "no ads" bet?
Good question. (cit.)
Kagi, the paid ad-free search engine, has been promising this for years: ten dollars a month, no ads, no tracking. They have 38,000 paying members. In a world of billions of Google users, "no ads" wasn't a compelling enough reason to switch.
But Anthropic isn't Kagi. Claude competes with ChatGPT on quality, not just principles. The question is whether Anthropic can keep that promise, or whether they'll succumb. Just as OpenAI did, fifteen months after Sam Altman called advertising a "last resort".
Update, September 2026: seven months later, Claude is still ad-free. And Perplexity, which had tested ads since 2024, dropped them because they hurt trust in its answers.
Google's Gemini ads: already happening in Search
Adweek reported that Google briefed advertisers about Gemini ads for 2026. Google's VP of Ads denied it. But Google Search already shows ads in AI Overviews. Make of that what you want.
Update, September 2026: at Google Marketing Live in May, Google announced two ad formats for AI Mode. Both are still in testing. The Gemini app itself still has no ads.
Of course, for Google this isn't new territory. When they launched AdWords in 2000, ads were the business model from day one. OpenAI did it the other way around: subscriptions first, then ads. Either way, the destination is the same.
What about the marketers?
The ad-supported AI model is now the default. Users will tolerate it. The question is what this means on the other side of the equation.
For marketers, this is a bigger story. ChatGPT may have just created the most valuable advertising channel since Google Search: one where the user has already told the AI exactly what they need, why, and how much they want to spend. We break down the intent advantage, and how pricing moved from $60 CPMs to pay-per-click, in ChatGPT ads solve the intent problem.
Will you pay €23 to avoid ChatGPT ads, or just accept them? Tell us on LinkedIn.