Google executives, while previously stating no immediate plans, are now not ruling out placing ads within the Gemini AI chatbot. The company is currently experimenting with ads in its AI-powered Search product (AI Mode) to inform future decisions, noting its strong financial position allows it to avoid rushing monetization.
The AI chatbot market is highly competitive, with Google's Gemini growing to over 750 million monthly users as it races to catch up with OpenAI. Both companies are now focused on monetizing free users, with OpenAI already testing ads in ChatGPT, while other AI firms like Anthropic and Perplexity express caution about the approach.
Key topics covered include the potential for ads in Gemini, the competitive landscape of AI chatbots, and the broader industry's exploration of monetization strategies. The article also touches on how AI is improving Google's core advertising business and the unresolved question of integrating personalized user data with ads.
Google executives have insisted for months that the company has no immediate plans to put ads in Gemini. But in an interview with WIRED, Google’s senior vice president of knowledge and information, Nick Fox, says the tech giant is “not ruling them out.”
“I would expect that the learnings that we get from ads in AI Mode would likely carry over to what we might want to do in the Gemini app down the road,” says Fox. “It’s an odd thing to say, but our research shows that users actually like ads within the context of Search. Over time, we’ll figure out what makes sense in the Gemini app.”
Google has spent the past year racing to catch up with OpenAI in the AI chatbot market. Its efforts appear to be paying off. Gemini now has more than 750 million monthly active users, compared to the 350 million it had in March of last year (OpenAI doesn’t release monthly active user numbers, but its weekly active user number is 900 million).
Now, the question for both companies is how to make money from free users. In January, OpenAI announced it would start testing ads on ChatGPT’s free tier in the United States. Naturally, this raised questions as to when Google—which runs the largest online advertising businesses—would follow suit.
Google DeepMind CEO Demis Hassabis tried to put a stop to that speculation, telling reporters at Davos the following week that the company didn’t have “any plans” to put ads in Gemini.
Instead, Google is testing ads in AI Mode, the Search product powered by Gemini.
“We have an environment with AI Mode where we can experiment [with ads],” Fox says. He notes that Google’s business is doing quite well these days—2025 was the company’s first year generating more than $400 billion in revenue—so it doesn’t have to rush to monetize Gemini. That puts Google at an advantage compared to OpenAI. The ChatGPT-maker reportedly aims to more than double its $30 billion revenue in 2026, with a significant chunk of the growth expected to come from ChatGPT.
Anthropic is taking the opposite route, running a Super Bowl commercial last month highlighting the potentially disastrous impact of ads in AI. This sparked a broader conversation around how the AI industry can do ads in a way that’s helpful and preserves people’s privacy. In February, Perplexity executives said they would stop experimenting with ads in its AI, partly because of the impact on user trust.
One open question for Google is how ads will interact with Personal Intelligence, a feature the company launched in January, which allows Gemini to reference a user’s Gmail, Photos, and Calendar to generate contextual responses. Fox says it’s “TBD” whether Personal Intelligence will make it into traditional Search but notes that personalizing Search has long been his “holy grail.”
This interview has been edited for clarity and length.
WIRED: Nick, there was a lot of concern a few years ago that AI was going to cannibalize Google, but the business has continued to do better than ever. How is AI changing Google’s ads business?
Nick Fox: On the ads side, there's three different pieces. First is improving core ad quality. For every ad, we generate a prediction of how well that ad will perform on a query. Will the user click on it? Is the user likely to convert? Turns out, AI models do a better job at those predictions. I would say that’s the biggest way Gemini is improving our ads.
Second is advertiser tools. If you’re a small business, you’re not thinking about all the queries people are going to type in. AI is great at figuring out which keywords to use, what’s the optimal creative, and generating all of that.
The third piece is the most nascent: ads in new experiences. The general philosophy we have is to build a great consumer product, then figure out monetization. Because the business is so strong and healthy, that's a luxury we have.
What have you learned from experiments around ads in AI Mode?
Ads are always separate from organic results and clearly labeled. If we don’t think any ad is relevant, we don’t show any ads. Probably the biggest principle of all is that ads should be useful.
What [ads in AI Mode] have shown is mostly intuitive things. If it’s relevant, a user will click on it. If not, they won’t.
At Davos, Demis Hassabis said that Google has no plans to bring ads to Gemini. How are you thinking about it now?
The reason we focused on ads in AI Mode and AI Overviews is because we see them as extensions of the Search experience. It’s the most natural place for us to do initial experiments.
I would expect that the learnings that we get from ads in AI Mode would likely carry over to what we might want to do in the Gemini app down the road. We're able to get all those learnings within a context and a construct that users are already aware of ads. It’s an odd thing to say, but our research shows that users actually like ads within the context of Search. Over time, we’ll figure out what makes sense in the Gemini app.
So you guys aren’t ruling out ads in Gemini completely?
No, we’re not ruling them out. It’s just not where we’ve been focusing.
Gemini is a massive product now—it’s grown quite fast. OpenAI is already trying out ads in ChatGPT. What makes you think Gemini isn't ready for ads?
We’re super happy with how well it’s growing. I would say it’s more of a prioritization question—what’s the right area to focus right now?
Do you think OpenAI introduced ads too early?
It’s hard for me to say. I think it will really depend on how they do it. What we’ve seen is when we do ads right, it’s accretive to the product experience. But the really critical thing is to do it right, and that means relevance, quality, and not putting ads where users don’t want to see ads.
That’s hard to do. We have over 20 years of experience learning how to do that. It’s less a question of timing and more a question of doing it right, and in a way that’s respectful of users.
Google recently launched Personal Intelligence in Gemini and AI Mode. I imagine advertisers would love to get their hands on that data. How are you thinking about that?
Personal Intelligence is incredibly useful. For example, I was skiing and couldn't see well through my goggles—it was cloudy out. I asked AI Mode a fairly vague question about what lens I needed for the conditions. From my email, it was able to say which resort I was staying at, which mountain I'd be skiing, and what the weather would be. It also pulled in a receipt my wife had forwarded me for my goggles—turns out she'd bought me an extra lens. It’s like subtle magic.
From an advertising point of view, there are no ads in the Gemini app at this point. Within AI Mode, we’re still working through it. We don’t sell data to advertisers. All this stuff is very nascent, but probably what we’ll figure out is how to make sure that an ad can be targeted and relevant to the context in a query. Maybe in this case, it could have known the specific brand of goggles.
We’ll need to figure out how the targeting of the ads can be consistent with the organic response and to make sure it all fits together. But we’ll make sure that your private information remains private and isn’t shared with advertisers.
I’m guessing you saw Anthropic’s Super Bowl commercial about ads in AI. What was your reaction to it?
I thought the ads were funny. From our point of view, ads in Search have worked incredibly well for 20-plus years. It doesn’t change our confidence that we can build incredibly useful, relevant ads in the context of AI search.
Search has a lot of AI in it today. Where is it heading?
The broad vision for Search is you should be able to ask anything. If you can ask all the questions in your head, you’ll probably visit more web pages, and you might buy more products.
I would say that personalization, for years, has been the holy grail of Search, but we never actually figured it out. For me, seeing the rollout of Personal Intelligence has been really game-changing in terms of Search getting more useful. You should expect to see us push quite a bit on that.
Do you mean Personal Intelligence will be a bigger part of AI Mode, or Search in general?
So far, we’ve brought it to AI Mode. We think of AI Mode and AI Overviews as increasingly one and the same. A core part of how we’re evolving Search is that it should be more seamless. Users should just bring whatever question to Search and we should show you the most useful response. That might be a navigational link to a webpage. It might be a set of news articles. It might be an AI response.
All those things come together, and that’s a reasonable way to think about how Personal Intelligence will evolve as well.
Do you expect all of these products—AI Mode, AI Overviews, Gemini, and Search—to be one thing at some point?
It kind of depends on how usage evolves. Today, the way we think about it is Search is becoming AI search, and the Gemini app is your personal assistant. Those areas have overlap for sure.
Within the context of Search, users shouldn’t need to be thinking quite as much about what’s traditional Search, what’s AI Overviews, what’s AI Mode— that really should come together into a seamless, holistic Search experience. That’s the direction we’re going.
If Personal Intelligence gets wrapped up into Search more, that means that people's searches could be influenced by their emails, photos, and all of these different things. How are you guys preparing for that shift?
One thing we’ve been clear about with Personal Intelligence is that it’s opt-in. We don’t want people to be surprised. We think Personal Intelligence is super helpful and would love lots of users to use it, but if users don’t feel comfortable, they shouldn’t need to. This is not something that's being done to users, but this is something that users are choosing to invoke.
In terms of how [Personal Intelligence] connects back to traditional Search, I think that’s TBD—we wouldn’t do that in a way that’s uncomfortable for users, and we may never do it—but it’s based on how the technology evolves, how the use cases evolve, and whether users want it.
Meridiano
