White House Meets With Anthropic Over New AI Model With Cybersecurity Implications
White House Chief of Staff Susie Wiles met Friday with Anthropic CEO Dario Amodei to discuss the company's new Mythos model, which the federal government views as potentially significant for national security and the economy.
A White House official said the administration is evaluating advanced AI models and their security implications. Any technology the federal government uses would require a technical evaluation period before deployment.
The White House called the meeting productive and constructive. Both sides discussed collaboration opportunities and balancing innovation with safety.
Anthropic said the conversation covered cybersecurity, U.S. competitiveness in AI, and AI safety. The company indicated it plans to continue talks with the administration.
Context: Recent Tensions Between Trump Administration and Anthropic
The meeting signals a shift in tone. Months earlier, President Trump ordered all federal agencies to stop using Anthropic's Claude chatbot over a Pentagon contract dispute, posting on social media that the administration "will not do business with them again!"
Defense Secretary Pete Hegseth also sought to declare Anthropic a supply chain risk-an unprecedented action against a U.S. company. Anthropic challenged both moves in federal court.
The company had sought Pentagon assurances that its technology would not be used in fully autonomous weapons or surveillance of Americans. Hegseth rejected those conditions.
A federal judge blocked Trump's directive in March, allowing federal agencies to continue using Anthropic products.
What Mythos Does
Anthropic announced Mythos on April 7, describing it as exceptionally capable at finding and exploiting computer vulnerabilities-work that typically requires human cybersecurity experts. The company is limiting access to select customers due to these capabilities.
The UK's AI Security Institute evaluated the model and called it a "step up" over previous versions. The institute said Mythos can exploit systems with weak security defenses and predicted other companies will develop similar models.
Even skeptics of Anthropic's safety claims acknowledge the model represents real progress. David Sacks, the White House's AI and crypto tsar and a vocal Anthropic critic, said the cybersecurity concerns "make sense" as coding models improve. "As the coding models become more and more capable, they are more capable at finding bugs," Sacks said on his podcast. "That means they're more capable at finding vulnerabilities."
Government Response and International Interest
The federal government is not alone in paying attention. The European Union is in talks with Anthropic about the company's AI models, including advanced versions not yet released in Europe.
Anthropic formed an initiative called Project Glasswing to address the model's risks. The effort includes Amazon, Apple, Google, Microsoft, and JPMorgan Chase, among others, working to secure critical software from potential harm.
Anthropic co-founder Jack Clark said at a conference this week that Mythos, while advanced, is not unique. "There will be other systems just like this in a few months from other companies, and in a year to a year-and-a-half later, there will be open-weight models from China that have these capabilities," Clark said. "So the world is going to have to get ready for more powerful systems."
For government professionals, the meeting underscores an emerging reality: federal agencies will need to evaluate and potentially adopt advanced AI tools while managing the security risks they present. See more on AI for Government and AI for Cybersecurity Analysts.
Your membership also unlocks:
