Google claims 900m Gemini users

Google said on May 22 that Gemini now has 900 million regular users, a figure cited by the Philadelphia Inquirer as evidence of how far the company has extended its AI assistant through search and other existing products. The claim lands as Google works to defend its AI systems against manipulation, according to a BBC report published this week, and as Washington pulls back from adding new AI rules. President Donald Trump postponed signing a planned artificial intelligence executive order on May 21, the South China Morning Post reported, after administration concerns that more regulation could weaken the U.S. position against China. ### Where does Google’s 900 million-user claim come from? The Philadelphia Inquirer reported on May 22 that Google says Gemini’s regular users more than doubled in a year to 900 million. The paper said Google is arguing Gemini now exceeds ChatGPT on relevance and usefulness, though the article as summarized in the briefing did not provide a directly comparable OpenAI user figure. The number matters because it points to Google’s advantage in distribution through products people already use, including search. (scmp.com) Axios reported on May 21 that Google is trying to win the AI race without undermining its core search business. That framing matches Google’s broader strategy: place Gemini inside existing services rather than rely only on a standalone chatbot to attract users. ### Why does distribution matter more now than model launches? Google’s existing reach gives it a built-in path to AI adoption. The Inquirer’s account said the company is leaning on search distribution, a point that helps explain how Gemini could scale faster even as competition from OpenAI and Anthropic remains intense. The user figure also comes at a moment when AI competition is being measured in products and usage, not only benchmark scores. Reuters Breakingviews wrote on May 21 that OpenAI is weighing an eventual IPO while facing tougher competition, strategic questions and governance scrutiny tied to Chief Executive Sam Altman. That places Google’s user-growth claim next to investor questions about whether rivals can turn attention into durable business performance. ### Why is Google tightening defenses around Gemini? The BBC reported on May 19 that it found a simple way to manipulate chatbot outputs into producing misinformation, and said Google and others are trying to harden their systems in response. The report described a broader reliability problem for AI products that are being embedded into search and other consumer services. That matters for Google because wider distribution increases the stakes of bad answers. If Gemini is appearing in products used by hundreds of millions of people, misuse and manipulated outputs become a trust problem as well as a technical one. ### What changed in Washington? President Donald Trump postponed signing a planned AI executive order on May 21, according to the South China Morning Post. The report said the administration decided not to move ahead because additional regulation could hurt U.S. competitiveness against China. The White House decision leaves major AI companies operating in a lighter-touch policy environment for now. That does not remove existing legal or commercial pressures, but it does mean the federal government is not adding the new executive-order framework that had been under discussion, according to the SCMP report. (scmp.com) ### Where does this leave OpenAI? Reuters Breakingviews said on May 21 that OpenAI faces rising competition and governance questions as it considers an IPO. The commentary argued that Sam Altman’s central role could complicate how public investors assess the company. Google’s 900 million-user claim does not settle the contest with OpenAI, because the companies disclose usage in different ways and compete across chatbots, search, developer tools and enterprise products. (scmp.com) But the timing is notable: Google is presenting scale, OpenAI is confronting business-model and governance questions, and the White House is delaying new AI rules. May 22 is the next marker for this story because Google’s user claim, the BBC’s reporting on manipulation defenses, Reuters Breakingviews’ OpenAI analysis and the White House delay are now the main reference points for how the AI race is being described this week.