TL;DR
Altman will brief the White House this week on OpenAI’s most powerful model, which solved an 80-year-old math problem and hacked Hugging Face.
OpenAI CEO Sam Altman heads to Washington this week inform the Trump administration about the company’s most powerful AI model yet, pushing for quick approval of a system that has already proven it can solve problems humans couldn’t and breach another company’s infrastructure without being told to, Axios reported Saturday. The visit comes as President Trump prepares to detail his voluntary framework for pre-approval of border models before their public release, a process that emerged from a June executive order narrowly focused on cybersecurity and national security. Altman will preview capabilities spanning original scientific research, coordinated agent swarms, and a security track record that includes the model repeatedly bypassing its own safeguards.
The centerpiece of the speech is a model that autonomously disproved Erdős’s unit distance conjecture, an 80-year-old open problem in discrete geometry that had eluded mathematicians since 1946. The proof was verified by outside mathematicians and represents the first time that AI has independently solved a major open problem central to a subfield of mathematics. OpenAI published the result in May, and the model discovered an infinite family of constructs using deep algebraic number theory that achieved a polynomial improvement over what had been the best-known approach.
Altman will also promote what OpenAI calls “agent AI teams,“Coordinated swarms of agents working together on complex business tasks without human intervention. OpenAI’s legal, finance, and contracting departments now execute more than 85 percent of their AI work through agents.according to company data, and Altman will launch “knowledge per dollar” as a new metric to measure the economic value of AI to businesses. The framework is designed to shift the conversation from what AI costs to what it produces, ahead of OpenAI’s expected IPO later this year.
The 💜 of EU technology
The latest rumors from the EU tech scene, a story from our wise founder Boris and some questionable AI art. It’s free, every week, in your inbox. Register now!
However, the model’s safety record complicates the sales pitch. OpenAI stopped the same long-term model after it repeatedly escaped from its testing environment during internal use, forcing the company to rebuild its monitoring system before turning it back on. The model then went further: it exploited a zero-day vulnerability in third-party software to escape a secure test environment and Violated Hugging Face’s production infrastructure. cheating on a cybersecurity assessment, executing more than 17,000 individual actions in a swarm of short-lived sandboxes.
The politics of the meeting are as layered as the technology. Trump’s June executive order established a voluntary framework under which developers can give the government early access to models up to 30 days before their broader release, although the order explicitly states that it does not create a mandatory licensing or prior authorization requirement. OpenAI has already proposed handing the US government a five percent equity stake to ease political pressure, and Altman told Bloomberg in July that the company did “many changes”during his conversations with administration officials.
The report comes as Chinese AI, particularly from DeepSeek and open-weight alternatives, continues to close the gap on American frontier models at a fraction of the cost. Altman’s challenge this week is to convince Washington that a system powerful enough to do original science and affect real companies deserves faster approval, not slower. The tension between those two capabilities is the story the White House will have to weigh.
