Anthropic's CEO was expected to visit President Donald Trump's chief of staff, Susie Wiles, on Friday to discuss the artificial intelligence company's dispute with the War Department, according to reports.

Dario Amodei's White House meeting marks a potential turning point in a high-stakes standoff between the fast-rising AI firm and the Pentagon, which recently labeled Anthropic a national security supply-chain risk and effectively barred its technology from military use.

The Trump administration is weighing the potential of Anthropic's latest AI model, known as "Claude Mythos," against concerns about its powerful and potentially dangerous cybersecurity capabilities, Axios reported Friday.

Sources close to the discussions say officials recognize the risks of sidelining such cutting-edge technology.

"It would be grossly irresponsible for the U.S. government to deprive itself of the technological leaps that the new model presents," one source said, according to Axios, warning it could hand a strategic advantage to China.

The dispute escalated after Anthropic sued the Pentagon for blacklisting the company after it refused to allow unrestricted use of its AI systems.

Pentagon officials have taken a hard line, with Chief Technology Officer Emil Michael saying there was "no chance" of negotiations following what he described as bad-faith dealings.

Still, Friday's meeting suggests the White House may be seeking a broader resolution, particularly as parts of the U.S. government, including intelligence agencies and the Cybersecurity and Infrastructure Security Agency, already are testing the technology.

The urgency surrounding the talks reflects growing alarm in Washington over the rapid evolution of artificial intelligence.

Anthropic's Mythos model has demonstrated an ability to uncover vulnerabilities in critical systems, raising both defensive opportunities and national security risks.

Experts warn that such tools could lower the barrier for sophisticated cyberattacks, potentially allowing hostile nations such as China, Iran, or Russia — or even smaller actors — to exploit weaknesses in U.S. infrastructure.

At the same time, the administration is under pressure to ensure the United States does not fall behind in the global AI race.

Conservatives have increasingly emphasized the need to maintain technological dominance over geopolitical rivals, particularly China, which has invested heavily in advanced computing and cyberwarfare capabilities.

Anthropic has sought to position itself as a leader in responsible AI development, placing limits on how its most powerful models can be used.

The company has restricted access to Mythos to a small group of major corporations and launched a cybersecurity initiative aimed at identifying and fixing vulnerabilities before they are exploited.

Behind the scenes, Axios reported that Anthropic has also enlisted advisers with close ties to Trumpworld, signaling its willingness to find common ground with the administration after months of legal and political conflict.

Reuters contributed to this report.

• British Regulators Assess Risks of Anthropic's Latest AI Model
• Scott Bessent Warns Banks on AI Cyber Threat