Google Opens the Door to Ads in Gemini After Months of Denials

Google Reverses Course on Gemini Advertising

Google has officially shifted its position on placing advertisements in its Gemini AI assistant. In an interview published on March twelfth, Nick Fox, Google's senior vice president of knowledge and information, told WIRED that the company is "not ruling out" advertising in the chatbot, a notable departure from categorical denials made just weeks earlier.

A Pattern of Denials

The reversal follows at least two public assurances from senior Google executives. In December twenty twenty-five, Google Ads VP Dan Taylor said there were no plans to change Gemini's ad-free status. Then in January, DeepMind CEO Demis Hassabis told an audience at the World Economic Forum in Davos that Google had "no plans" for Gemini ads, even questioning why competitor OpenAI had moved so quickly to monetise ChatGPT.

AI Mode as a Testing Ground

Fox revealed that Google is already experimenting with ads in AI Mode, its Gemini-powered search product, and treating it as a proving ground. He indicated that learnings from those experiments would "likely carry over" to the standalone Gemini app. Google has outlined a three-part framework for AI advertising: improving core ad quality predictions, building AI-powered creative tools for advertisers, and testing ads in new AI experiences.

The Scale of the Opportunity

The commercial incentive is substantial. Gemini now has over seven hundred and fifty million monthly active users, more than doubling from three hundred and fifty million a year ago. Alphabet reported surpassing four hundred billion dollars in annual revenue for the first time, with advertising remaining its primary revenue engine.

Competitive Pressures

The move comes as OpenAI launched ads in ChatGPT's free tier in February at a sixty dollar CPM with a two hundred thousand dollar minimum commitment. Internal OpenAI documents project that free-user monetisation could generate one billion dollars in twenty twenty-six, scaling to twenty-five billion by twenty twenty-nine. The economics of running large AI models remain punishing, with OpenAI's cumulative losses exceeding thirteen point five billion dollars in the first half of twenty twenty-five.

Trust at Stake

The central tension remains whether users will accept advertising in a personal AI assistant that may have access to their emails, photos, and calendars. Fox acknowledged this remains an open question, while Hassabis had previously warned that poorly executed ads could "erode user trust."