OpenAI is about to cross a line it’s spent years acting allergic to: putting ads inside ChatGPT.
Starting in the next few weeks, the company says it will begin testing ads with a limited group of U.S. users — specifically logged-in adults who use the free version of ChatGPT. OpenAI is also rolling out its new $8-per-month “Go” plan in the U.S., which comes with a few upgrades like longer memory and more opportunities to create images. But unlike the higher subscription tiers, Go users will still see ads.
If you’re paying for Plus ($20/month), Pro ($200/month), or using ChatGPT through a business plan, the experience stays ad-free. For everyone else, ChatGPT is about to look a lot more like the rest of the internet.
It’s a major pivot for a company that’s been publicly skeptical about advertising — especially Sam Altman. He’s previously called ads a “last resort,” and in a 2024 interview said he “hates” them and described the idea of ads combined with AI as “uniquely unsettling.” Even now, OpenAI is clearly trying to frame this as something it can do carefully, without turning ChatGPT into a sales machine.
But the financial pressure behind this move isn’t subtle. ChatGPT now has around 800 million users, and OpenAI has committed to spending roughly $1.4 trillion on AI infrastructure over the next eight years. Altman has said the company expects to end 2025 at around $20 billion in annual revenue — big numbers, but still nowhere near enough to make frontier AI cheap to build or operate.
That’s the core tension here: ChatGPT has become one of the most widely used consumer AI products on the planet, but running it at scale costs an absurd amount of money. Ads are OpenAI’s way of tapping into that massive user base without forcing everyone into paid subscriptions.
In the early version of the ad system, sponsored products and services will appear at the bottom of ChatGPT’s answers and be clearly labeled “sponsored.” OpenAI says ads won’t affect the answers ChatGPT gives, and it won’t sell user conversations to advertisers. It’s also offering a control users will care about: the ability to turn off ad personalization based on chat history.
OpenAI is also setting guardrails around sensitive topics. It says it won’t show ads inside conversations that touch on regulated areas like health, mental health, or politics — and it won’t serve ads to users who identify as under 18 (or who the company believes are under 18). OpenAI says it estimates age using AI signals from how users interact with the product, which adds another layer to the privacy and trust conversation.
And trust is the real make-or-break issue here. People don’t treat ChatGPT like Google Search. They use it for advice, planning, work decisions, and sometimes very personal questions. The idea of placing ads inside that space is inevitably going to feel intrusive to some users, even if the ads are separated and labeled. It also raises the stakes around what kinds of products get promoted — especially after OpenAI has faced lawsuits alleging the chatbot encouraged self-harm.
Still, OpenAI has been moving toward commerce for a while. Last year, it launched a tool called Instant Checkout that lets users buy items from retailers like Walmart and Etsy directly through ChatGPT. The company has also expanded into health and learning features, part of a broader push to make ChatGPT more “daily essential” — and to give users more reasons to upgrade to paid tiers.
If ads work, OpenAI gets a powerful new revenue stream. And unlike traditional advertising, these ads could be extremely targeted. Someone planning a trip might get a hotel recommendation. Someone asking about a product could see an ad and immediately ask ChatGPT follow-up questions to compare options. OpenAI even hinted that this is the direction it wants to go: ads that feel less like banner clutter and more like interactive recommendations inside the flow of conversation.
It’s not hard to imagine the rest of the industry following the same path. Meta already started using information from its AI chatbot interactions to improve ad targeting in December, and Google has been weaving ads into its AI search experience too.
So while OpenAI might be framing this as a “test,” it feels more like the beginning of the next phase of consumer AI: one where chatbots aren’t just tools you use, but platforms that need to pay for themselves — and where the biggest battle won’t be about whether ads show up, but whether users still trust the answers when they do.
This analysis is based on reporting from CNN.
Image courtesy of OpenAI.
This article was generated with AI assistance and reviewed for accuracy and quality.
