The Trump administration met with Anthropic CEO Dario Amodei on Friday to discuss the company's powerful new Mythos AI model, marking the first high-level engagement since a bitter dispute earlier this year over military contracts and ethical guardrails.
Chief of Staff Susie Wiles and Treasury Secretary Scott Bessent hosted Amodei at the White House as governments worldwide scramble to understand Mythos's unprecedented ability to identify and exploit cybersecurity vulnerabilities. The model, announced April 7, has been restricted to select customers due to its potential to supercharge cyberattacks.
We discussed opportunities for collaboration, as well as shared approaches and protocols to address the challenges associated with scaling this technology
White House statement
The meeting signals a potential thaw in relations after President Trump banned federal agencies from using Anthropic's Claude chatbot in February, declaring the administration would not do business with the safety-focused AI company again. Defense Secretary Pete Hegseth went further, attempting to designate Anthropic as a supply chain risk — an unprecedented move against a domestic company.
The conflict erupted when Anthropic demanded guarantees that its technology would not be used for autonomous weapons or domestic surveillance. The Pentagon rejected these conditions, insisting the company must allow any lawful government use. A federal judge blocked enforcement of Trump's directive in March, allowing the legal battle to continue.
Frames the story as a diplomatic reconciliation between pragmatic actors, emphasizing the global nature of cybersecurity threats and the need for international coordination. Downplays ideological tensions in favor of technical cooperation imperatives.
Emphasizes the contradiction between Trump's authoritarian tendencies and Anthropic's ethical stance, framing the dispute as emblematic of broader tensions over AI governance. Highlights European concerns about unchecked AI development and surveillance overreach.
Focuses on the procedural and legal aspects of the dispute while treating both sides' positions as legitimate policy differences. Emphasizes the national security implications and economic competitiveness concerns driving the reconciliation effort.
Banking officials in the United States, Canada, and Britain have held urgent meetings about Mythos's threat to financial institutions, which rely on technology systems mixing cutting-edge tools with decades-old software. The model's advanced coding capabilities give it what experts describe as an unparalleled ability to find and exploit system weaknesses.
Anthropic has deployed Mythos through "Project Glasswing," a controlled program allowing select companies including Google, Apple, and Microsoft to test the model's cybersecurity applications. The company describes Mythos as its "most capable yet for coding and agentic tasks," referring to its ability to operate autonomously.
Anytime Anthropic is scaring people, you have to ask, 'Is this a tactic? Is this part of their Chicken Little routine? Or is it real?'
David Sacks, former White House AI and crypto czar — All-In podcast
Even critics acknowledge Mythos may represent a genuine breakthrough. David Sacks, who previously served as Trump's AI czar and has been skeptical of Anthropic's safety claims, urged people to take the company's warnings seriously this time.
The White House described Friday's discussion as "productive and constructive," while Anthropic said it explored collaboration on cybersecurity, America's AI leadership, and safety protocols. Both sides indicated they would continue the dialogue with other leading AI companies.