Briefly: Anthropic CEO Dario Amodei met with White House Chief of Staff Susie Wiles and Treasury Secretary Scott Bessent on Friday for what the White House said were “productive and constructive” talks about access to Mythos, a frontier artificial intelligence model that can find thousands of zero-day vulnerabilities. The meeting marks a thaw in the standoff that began when the Pentagon blacklisted Anthropic for refusing to lift security restrictions, though any deal would exclude access to Mythos through the Defense Department and civilian agencies.
Anthropic CEO Dario Amodei headed to the West Wing on Friday to meet with White House Chief of Staff Susie Wiles and Treasury Secretary Scott Bessent. The White House called the conversation “introductory, productive and constructive,” saying the three discussed “opportunities for cooperation, as well as shared approaches and protocols to address the challenges of expanding this technology.” President Trump later told reporters that he had “no idea” that the meeting had taken place.
The meeting is the most important step towards resolving a standoff that has seen one of the world’s most important AI companies blacklisted by its own government, while the same government fights to gain access to its most powerful model. If the two sides come to an agreement, it would likely exclude the Pentagon entirely and channel access to Mythos through civilian agencies that were not parties to the original dispute.
How did we get here?
The conflict began in late February when Defense Secretary Pete Hegseth demanded that Anthropic provide unfettered access to the Pentagon’s AI models for “all legitimate purposes,” including autonomous weapons systems and internal surveillance. Amodei refused. He said publicly that Anthropic wants to work with the military, but that AI models are not yet reliable enough for autonomous weapons and that US laws do not protect Americans from using AI in mass surveillance. Hexeth’s response was to label Anthropic a national security supply chain risk, a classification previously reserved for companies with ties to foreign adversaries, effectively blacklisting it from all government contracts.
Anthropic sued the Trump administration in early March, filing two federal lawsuits alleging wrongful retaliation. A federal judge initially blocked the blacklist, but an appeals court overturned that decision on April 8. Anthropic is now excluded from Department of Defense contracts, but may still work with other government agencies. After the court ruling went against it, Anthropic hired Trumpworld advisers to facilitate a political settlement, and Axios reported Friday that the meeting was set up to pave the way for a settlement.
💜 of EU technology
The latest rumblings from the EU tech scene, a story from our wise founder Boris and some questionable AI art. Free in your inbox every week. Register now!
The paradox that brought Amodei to the White House is that Anthropic announced Mythos on April 7, ten days after it lost its appeal, and the model became something the government could not ignore.
What the Mythos can do
Mythos is a general-purpose artificial intelligence model that can identify and exploit thousands of previously unknown zero-day vulnerabilities in every major operating system and web browser during testing. It found flaws that had survived decades of human security testing. When an employee was directed to develop exploits, he succeeded in more than 83% of the first attempts. It is the first AI model to complete a 32-step enterprise network attack simulation from start to finish. The UK’s Artificial Intelligence Security Institute rated it as “more capable of cyberattacks than any model previously assessed”. Jamie Dimon, CEO of JPMorgan Chase, said publicly that this “reveals greater vulnerability” to cyberattacks. The Council on Foreign Relations called it a “tipping point for AI and global security.”
Anthropic chose not to release Mythos to the public. Instead, he created Project Glasswing, a managed access program that provides a model for finding and fixing vulnerabilities before they are exploited to about 40 vetted organizations, including Amazon Web Services, Apple, Google, Microsoft, Nvidia and JPMorgan Chase. The company committed up to $100 million in Mythos usage credits and $4 million to open source security organizations. The decision to restrict rather than release is a direct application of the security principles that put Anthropic at odds with the Pentagon in the first place.
What each side wants
The Treasury Department is looking to Mythos to hunt for vulnerabilities in its systems. Part of the intelligence community and the Cybersecurity and Infrastructure Security Agency are already testing it. The White House Office of Management and Budget is building safeguards to allow federal agencies to use the controlled version. Bessent’s presence at Friday’s meeting suggests that the economic and financial security arguments for access to Mythos are reaching the highest levels of the administration.
Anthropic blacklisting needs a solution. Not because he needs the Pentagon revenue; the annual revenue of the company It has reached $30 billion, attracted $800 billion worth of investor bids, and is exploring an IPO. But supply chain risk assignment is hurting it reliability of the enterprise and creates uncertainty for every customer adjacent to the government. What Amodei wants is a resolution that restores his company’s position without surrendering its controversial security obligations.
The outlines of a compromise appear. Anthropic will provide access to Mythos for defense cybersecurity purposes through civilian agencies. Management will withdraw or narrow its supply chain risk designation. Unless a separate process is agreed upon to review special military use cases, the Pentagon will remain overseas. Both sides have incentives: anthropic because the blacklist is commercially harmful, and the White House because the technology is abundant worth giving up.
Pressures from abroad
The diplomatic dimension adds urgency. Anthropic plans to introduce Mythos to choose from British banks during the day and quadrupling its London office to 800 employees. Bank of England Governor Andrew Bailey called Mythos a cyber security risk in a speech at Columbia University on April 15, and the Bank’s Cross Market Operational Resilience Team is calling an urgent briefing with the CEOs of the UK’s eight largest banks and representatives from the Treasury, CCA, Homeland Security and CCA. Canadian Finance Minister François-Philippe Champagne described Mythos as an “unknown unknown” at IMF meetings.
The result is a situation where America’s closest allies may gain access to a critical national security tool before the US government. This geopolitical reality It gives the White House an incentive to resolve the dispute, which goes beyond the initial disagreement over security guards. Bessent, one of the agencies most seeking access to the Treasury Department’s Mythos, likely made that point during Friday’s meeting.
What does Friday mean?
The word “introduction” is chosen carefully in the White House reading. This suggests that Wiles and Bessent opened a channel, not a deal. The trial is still ongoing. The appellate court ruling still stands. Hegseth did not back down from his position. But the White House Chief of Staff and Treasury Secretary sitting down with the CEO of a Pentagon-blacklisted company and calling the conversation productive represents a shift in the administration’s stance that was hard to imagine six weeks ago.
Amodei built the most capable cybersecurity tool available as a byproduct of building a general-purpose AI model, then limited its release for security reasons, then got punished by a government for maintaining the same security principles, and is now controlled by that government because the tool can’t be replicated or modified. This sequence plays out not in a congressional hearing or a regulatory process, but in a room in the West Wing where the most powerful chief of staff in a generation, the Treasury Secretary, and the CEO of an AI company try to find a formula that satisfies national security, commercial reality, and the security principles that started the whole fight. Friday did not create this formula. But he found that everyone in the room wanted him.
