White House Chief of Staff to Meet Anthropic CEO About New Mythos AI Model
White House Chief of Staff Susie Wiles will meet Anthropic CEO Dario Amodei at the White House to discuss Anthropic's new Mythos AI model. The meeting is meant to address security concerns around the model and to discuss possible federal use under a technical evaluation period.
Anthropic unveiled Mythos on April 7 and called it "strikingly capable," restricting access because the company says the model can outpace human cybersecurity experts at finding and exploiting software flaws. A White House official said the administration is engaging advanced AI labs about both their models and software security. Critics such as David Sacks said on the All-In podcast that people should "take this seriously" and credited Mythos' cyber skills as a real risk rather than a marketing scare.
The meeting comes amid a wider dispute with the Pentagon. Defense Secretary Pete Hegseth sought to label Anthropic a supply chain risk after a contract fight over bans on fully autonomous weapons and U.S. domestic surveillance use. Anthropic is fighting the designation in two federal courts. In March a U.S. district judge blocked enforcement of President Trump's social media directive that ordered federal agencies to stop using Anthropic products.
Coverage has shifted from punitive steps to cautious engagement. Earlier reports focused on bans, a blocked enforcement order, and the Pentagon's supply chain move. Newer reporting emphasizes that the White House wants technical talks and evaluations, with PBS breaking news of the scheduled meeting.
đ Key Facts
- White House chief of staff Susie Wiles will meet Anthropic CEO Dario Amodei at the White House on Friday to discuss Anthropic's new Mythos AI model.
- A White House official says the administration is engaging advanced AI labs about their models and software security, and that any federal use of such models would require a technical evaluation period.
- Anthropic announced the Mythos model on April 7, describing it as "strikingly capable" and limiting access because it can surpass human cybersecurity experts in finding and exploiting software vulnerabilities.
- Former White House AI and crypto czar David Sacks warned people to "take this seriously," saying Mythos' likely cybersecurity capabilities are real and explaining that as coding models grow more capable they naturally become better at finding bugs, chaining vulnerabilities and creating exploits.
- Defense Secretary Pete Hegseth sought to designate Anthropic a Pentagon supply-chain risk after a contract dispute over banning fully autonomous weapons and U.S. domestic surveillance use; Anthropic is challenging that move in two federal courts.
- In March, U.S. District Judge Rita Lin blocked enforcement of President Trump's social media directive that ordered all federal agencies to stop using Anthropic products.
đ° Source Timeline (2)
Follow how coverage of this story developed over time
- White House chief of staff Susie Wiles will meet Anthropic CEO Dario Amodei at the White House on Friday to discuss Anthropic's new Mythos AI model.
- A White House official says the administration is engaging advanced AI labs about both their models and software security, and stresses any federal use would require a technical evaluation period.
- Anthropic announced its Mythos model on April 7, describing it as 'strikingly capable' and limiting access because it can surpass human cybersecurity experts in finding and exploiting software vulnerabilities.
- Defense Secretary Pete Hegseth sought to designate Anthropic a Pentagon supply chain risk after a contract dispute over banning fully autonomous weapons and U.S. domestic surveillance use, and Anthropic is challenging that move in two federal courts.
- U.S. District Judge Rita Lin in March blocked enforcement of President Trump's social media directive ordering all federal agencies to stop using Anthropic products.
- Former White House AI and crypto czar David Sacks, a prominent Anthropic critic, said on his 'All-In' podcast that people should 'take this seriously' and credited Mythos' likely cybersecurity capabilities as 'more on the real side' rather than a marketing scare tactic.
- Sacks explained that as coding models become more capable they naturally become better at finding software bugs, chaining vulnerabilities and creating exploits, implicitly validating Mythos' described cyber risk.