Anthropic Engages with US Officials Amid Legal Challenges
Anthropic, a company known for its AI safety platform, is in a unique situation in Washington, D.C. While the company is currently in a legal battle with the U.S. government, it is also part of a select group of tech leaders participating in closed discussions with senior officials from the Trump administration, including Vice President JD Vance.
Dario Amodei, CEO of Anthropic, was among several key tech leaders who took part in a conference call last week with Vance and Treasury Secretary Scott Bessent. This meeting focused on the security risks associated with advanced AI models. According to CNBC, this conversation happened just before Anthropic released its new Mythos model to a limited number of companies, including major players like Microsoft, Google, and Apple.
In addition to Amodei, prominent figures on the call included Sam Altman from OpenAI, Sundar Pichai of Google, Satya Nadella from Microsoft, Elon Musk of xAI, and others. This comes as Anthropic prepares to launch its most powerful AI model yet, which will be limited to around 40 major technology firms due to concerns over cybersecurity risks.
Continuing Dialogues Despite Legal Issues
Amodei’s involvement in these discussions is noteworthy given Anthropic’s fraught relationship with the federal government. The Trump administration is attempting to remove Anthropic’s Claude platform from federal agency use. The company is embroiled in a legal dispute over a risk designation from the Department of War that has barred it from Pentagon contracts.
Although a federal judge in San Francisco granted a temporary injunction to prevent the blacklisting, a federal appeals court later denied a request to extend this block. Anthropic confirmed that it briefed “senior officials across the U.S. government on Mythos Preview’s full capabilities,” including both its potential risks and benefits.
A representative from the company emphasized the importance of keeping the government informed about the model’s capabilities before releasing it publicly, ensuring that all parties are aware of the risks involved and how they are being managed.
