White House Meets Anthropic CEO Over Mythos AI Fears
The White House reported a positive discussion with Anthropic's leader, even as the AI company pursues a lawsuit against the US Department of Defense. Officials described the session as productive and constructive. It occurred one week after Anthropic unveiled a preview of its Claude Mythos, an AI system the firm says surpasses human performance in certain hacking and cybersecurity activities.
Anthropic CEO Dario Amodei met on Friday with Treasury Secretary Scott Bessent and White House Chief of Staff Susie Wiles. Axios provided details on the encounter. An Anthropic spokesperson offered no remarks about it. The talks happened two months after White House officials called the company a radical left, woke outfit.
Capabilities of Claude Mythos
Access to Mythos remains limited to just a few dozen companies so far. Researchers note its strong skills in computer security work. The tool spots bugs in code from decades ago, Anthropic states, and it independently devises methods to exploit those flaws.
Last week, Amodei mentioned conversations with various US government officials. He proposed partnerships with them. Anthropic builds large language models under the Claude brand. The company started in 2021, led by Amodei and his sister Daniela, both former OpenAI staff. It prioritizes AI safety and alignment with human values. Mythos represents a step forward in specialized AI for security tasks.
Friday's dialogue signals the importance of Anthropic's tech to the government. This holds true despite the Trump team's hard line against the firm. The White House statement covered chances for joint efforts. It also addressed common methods and rules for handling growth in this technology. Officials examined how to advance progress while maintaining safeguards.
Background on the Legal Conflict
Stay ahead of the AI curve
The most important updates, news, and content — delivered weekly.
No spam. Unsubscribe anytime.
Anthropic filed a lawsuit in March against the Defense Department and other federal bodies. This followed its designation as a supply chain risk. Such a label marks a technology as insufficiently secure for government applications. No other US company had received this public notice before.
Anthropic has supported high-level government and military projects since 2024. In court, it claimed the label stemmed from retaliation by Defense Secretary Pete Hegseth. Amodei had denied the Pentagon unlimited access to its AI over concerns about domestic surveillance and autonomous weapons. A California federal court sided mostly with Anthropic. However, a federal appeals court rejected a temporary halt to the risk label.
Government agencies continue using Anthropic's tools as before the designation.
Administration's Prior Views and Recent Developments
Before Friday, the White House avoided favorable comments on Anthropic. President Trump ordered agencies to cease use of its products. On social media, he called the company run by left wing nut jobs trying to strong arm defense. He added, we don't need it, we don't want it, and will not do business with them again!
Reporters questioned Trump about Amodei's White House visit as he arrived at an event in Phoenix, Arizona, on Friday. The president replied he had no idea about the meeting.
This encounter highlights tensions between national security needs and political disputes in the AI sector. Anthropic's role in cybersecurity underscores its value, even amid legal battles.
Related on Neura Market