CEO of blacklisted Anthropic and White House hold ‘productive’ discussions on AI

White House and Anthropic’s CEO Engage in Strategic AI Talks Amid Legal Dispute

On Friday, Anthropic’s CEO Dario Amodei met with White House Chief of Staff Susie Wiles during a high-stakes session. The encounter occurred as his company faces ongoing legal challenges from the Trump administration over its exclusion from federal programs. Two sources close to the meeting confirmed the details to CNN, while Axios was the first to report on the interaction.

Amodei’s visit coincided with the government’s efforts to reconcile its strict stance on Anthropic with the strategic importance of its AI innovations. The firm’s technology, such as the Mythos tool, has dual capabilities: detecting cybersecurity threats and mapping paths for potential attacks on both corporate and governmental systems. The White House described the meeting as “introductory,” emphasizing its “productive and constructive” nature. In a statement, officials noted discussions focused on “collaboration opportunities” and “shared approaches to manage the risks of scaling this technology.” They added, “Balancing innovation with safety remains a priority, and we aim to keep this conversation open with other leading AI firms.”

“We discussed opportunities for collaboration, as well as shared approaches and protocols to address the challenges associated with scaling this technology,” the White House stated in a Friday afternoon release.

Anthropic similarly framed the dialogue as “productive,” highlighting topics like cybersecurity, the U.S.’s position in the AI race, and safety protocols. A spokesperson noted, “The meeting reflected Anthropic’s commitment to working with the government on responsible AI development. We are grateful for their time and eager to continue these discussions.”

Until recently, Anthropic’s Claude AI model was the sole tool accessible to the Pentagon’s classified network. However, Trump’s administration cut ties with the company after it resisted terms allowing military use of Claude for “all lawful purposes,” such as autonomous weapons and mass surveillance. The breakdown in negotiations led the Pentagon to label Anthropic a “supply chain risk,” a designation typically reserved for entities linked to foreign adversaries.

The company filed a lawsuit against the Trump administration, which was temporarily halted by a California federal judge. The judge ruled that federal agencies outside the Department of Defense couldn’t use the supply chain risk label to block access. The government has since appealed this decision. In a separate case, the DC Circuit Court of Appeals ruled in favor of the administration, permitting the Pentagon to sever ties with Anthropic during the legal process.

“would force the United States military to prolong its dealings with an unwanted vendor of critical AI services in the middle of a significant ongoing military conflict,” the appeals court stated in its unanimous decision.

Meanwhile, Anthropic unveiled a new AI model, Mythos, which it and experts predict could mark a pivotal shift in cybersecurity. The model offers early access to select groups for risk assessment, according to Bloomberg. The Office of Management and Budget has already informed agencies about plans to integrate Mythos into their systems, and the White House is reportedly exploring similar access for its own use.

Public records reveal Anthropic recently hired the lobbying firm Ballard, where Susie Wiles previously worked. The firm was specifically contracted to advocate for Anthropic’s procurement interests. A White House official noted, “The administration continues to engage across government and industry to safeguard the U.S. and its citizens. This includes working with frontier AI labs to ensure their models help secure critical software vulnerabilities.”