Sam Altman Proposes Global AI Forum as OpenAI Loses Ground to Anthropic and Google

Sam Altman published an op-ed in the Financial Times calling for an international forum, led by the United States, to set global safety standards for artificial intelligence. The piece ran at the turn of July 1-2 and quickly sparked debate over whether the proposal is a genuine attempt to improve AI safety or a bid to regain the initiative by a company that has been losing ground in the market in recent months.

Lessons from aviation and nuclear energy

Altman argues that AI will change the material conditions of human life on a scale comparable to the taming of electricity, perhaps greater. As reference points he cites global aviation safety standards, international financial regulation, and the International Atomic Energy Agency. The proposed forum would bring together government representatives and independent technical experts, establish recognized standards, provide unbiased assessments of models' capabilities and risks, and make the technology available to countries and companies that follow the shared rules.

The idea did not emerge in a vacuum. The proposal follows the G7 summit in France, where representatives of OpenAI, Anthropic and Google DeepMind met with world leaders to discuss shared standards for the most advanced models. According to reports, it was there that Altman first sketched out the idea of an international forum. He also notes that unlike aircraft or nuclear facilities, which inspectors can physically examine, AI models are built inside data centers with no outside visibility, making it far harder to verify whether a given lab is actually following the agreed rules.

Anthropic overtakes OpenAI

The business backdrop to the op-ed is telling. According to data cited by Deutsche Bank analyst Adrian Cox, Anthropic said in May that it is heading toward $47 billion in annualized revenue and expects to reach profitability by 2029, a year ahead of OpenAI's own plan. OpenAI's latest figures put its revenue at $25 billion to $33 billion on the same annualized basis.

Even more telling is the adoption data among businesses. According to May's Ramp AI Index, Anthropic's share of business usage rose 3.8 percentage points in April to 34.4 percent of companies, while OpenAI's share fell 2.9 points to 32.3 percent. It was the first time Anthropic overtook OpenAI in business subscriptions. Similarweb data also shows that ChatGPT's monthly visits fell below half of all traffic in the generative AI category for the first time in May, suggesting users are increasingly switching between models rather than sticking with a single provider.

A forum as narrative, not just policy

Some commentators note that Altman's proposal serves a dual purpose. On one hand, it is a genuine attempt to ease the competitive pressure pushing labs to release ever more powerful models faster without full control over the risks. On the other, it is also a way to reclaim the lead in the AI safety narrative at a moment when OpenAI is losing its market lead to Anthropic and Google. Positioning himself as the architect of global standards lets Altman stay at the center of the debate, even as the hard numbers on revenue and adoption tell a different story.

For the Polish market, this rivalry has concrete consequences. In recent weeks Anthropic rolled out Claude Sonnet 5 as the default model for free users, while Poland gained access to a special version of GPT-5.6 Cyber for NASK (Poland's national research and telecom institute) and CERT Polska (Poland's computer emergency response team). Companies and institutions choosing an AI provider today are operating in a market where the leader can change within a month, a fact that matters for long-term contracts and integrations.

It remains unclear whether Altman's proposal will win real political backing outside the United States, or whether Anthropic and Google DeepMind will even want to take part in a forum dominated by the US administration. Given that OpenAI's rivals are currently outperforming it in the market, it is hard to expect them to willingly hand the initiative to a competitor, even in the name of shared safety standards.

Sources: Sam Altman seeks new world order for AI as OpenAI slowly loses ground to Google and Anthropic (fortune.com), Sam Altman calls for US-led international forum to set global AI standards (siliconangle.com), Anthropic finally beat OpenAI in business AI adoption but 3 big threats could erase its lead (venturebeat.com).