ChatGPT is running ads now. Here's what it means for choosing AI tools.

OpenAI has begun putting advertising inside ChatGPT, and for anyone who relies on AI tools to find, compare, or buy software, this is the moment the ground shifted. The free chatbot that answered questions with no commercial layer is becoming an ad-supported product like the search engine it was supposed to replace. The details are still emerging and OpenAI has framed early formats as limited and clearly labeled, but the direction is set.

Why this was inevitable

ChatGPT reportedly serves hundreds of millions of weekly users, and inference is expensive. Subscriptions alone were never going to cover that bill at free-tier scale. Advertising is the obvious lever, and it is the same path Google took: give the answer away, then monetize the intent behind the question. The uncomfortable part is that a recommendation engine funded by advertisers has a structural incentive that a paid tool does not.

What changes for people who use AI to pick tools

A large and growing share of software discovery now happens through AI assistants rather than search. When you ask an assistant "what is the best AI voice generator" or "cheapest tool for X," you are trusting that the ranking reflects quality, not ad spend. Once advertising enters the loop, that trust needs qualifying. Three things to watch:

• Labeling. Sponsored suggestions must be visually distinct from organic ones. Early formats look labeled; the risk is drift over time as pressure to grow ad revenue builds.
• Placement bias. Even clearly labeled ads shape choices. The first option named in an answer carries disproportionate weight, and advertisers know it.
• Answer neutrality. The real question is whether ads stay in a separate slot or start influencing the body of the answer itself. The former is tolerable. The latter would quietly turn every product recommendation into paid placement.

The case for source-cited alternatives

This is where assistants that show their work gain an edge. Perplexity built its brand on citing sources for every claim, which makes it easier to check whether a recommendation is grounded in real evidence or is simply the confident-sounding default. That does not make it immune to a future ad model, but a cited answer is auditable in a way a bare one is not. If you are researching a purchase, an assistant that links its sources lets you verify before you buy.

What we are doing about it

We run this site on affiliate links, and we will not pretend otherwise: some links here earn us a commission. The difference we hold ourselves to is that the ranking comes first and the money second. Our editors rank tools on output quality, value, and ease of use, and the affiliate status of a tool never moves it up a list. You can read exactly how we pick on any of our best-of pages, where the reasoning behind each placement is on the page, not hidden in an algorithm.

The lesson from ChatGPT's advertising turn is not that AI recommendations are worthless. It is that you should know how any recommender is funded, and prefer the ones that tell you. Ask where the ranking comes from. If the answer is not clear, treat the top suggestion as a starting point, not a verdict.

The bottom line

Ads in ChatGPT are a milestone in the normalization of AI assistants as mainstream media. They will make the free tier more sustainable and, over time, more commercial. For everyday questions that is fine. For decisions where money changes hands, lean on tools that cite their sources, cross-check more than one assistant, and favor reviewers who disclose their incentives. The convenience of a single confident answer is exactly what makes an unlabeled ad inside it so effective.