The White Home stated its assembly with Anthropic CEO Dario Amodei was “productive and constructive” on Friday, the newest signal of a possible thaw within the standoff between the federal government and the main synthetic intelligence firm.
However when President Donald Trump was requested about Amodei’s go to on a runway in Phoenix, Ariz., he responded, “Who?” after which stated he had “no thought.”
Amodei met with Susie Wiles, the White Home chief of workers, and different senior administration officers to speak in regards to the firm’s highly effective new Mythos mannequin, which was introduced earlier this month. U.S. Treasury Secretary Scott Bessent was one of many officers in attendance, in line with a supply conversant in the matter who requested to not be named as a result of the assembly was personal.
“We mentioned alternatives for collaboration, in addition to shared approaches and protocols to deal with the challenges related to scaling this know-how,” the White Home stated in a press release on Friday. “The dialog additionally explored the stability between advancing innovation and making certain security.”
The corporate seems to be inching again into the White Home‘s good graces, weeks after Trump blacklisted it and known as it a nationwide safety danger, declaring that his administration would “not do enterprise with them once more.”
An Anthropic spokesperson stated Amodei had a “productive dialogue” with administration officers on Friday about how the corporate and the U.S. authorities can “work collectively on key shared priorities similar to cybersecurity, America’s lead within the AI race and AI security.”
Anthropic CEO Dario Amodei arrives for conferences with U.S. President Donald Trump’s admnistration officers on the White Home in Washington, D.C., U.S., April 17, 2026.
Jessica Koscielniak | Reuters
“The assembly mirrored Anthropic’s ongoing dedication to partaking with the U.S. authorities on the event of accountable AI,” the spokesperson stated. “We’re grateful for his or her time and are trying ahead to persevering with these discussions.”
Mythos excels at figuring out weaknesses and safety flaws inside software program, in line with Anthropic, and the corporate stated it has been partaking in “ongoing discussions” with U.S. authorities officers about its capabilities. Anthropic is rolling the mannequin out to a choose group of corporations as a part of Venture Glasswing, a brand new cybersecurity initiative, and doesn’t have plans to launch it publicly.
Anthropic declined to remark. Axios was first to report the assembly on Friday.
Simply weeks prior, Anthropic and the Trump administration have been firmly on the outs. Anthropic sued the federal government to attempt to reverse its blacklisting in courts in San Francisco and Washington, D.C., and people circumstances are ongoing.
The lawsuits adopted a dramatic few weeks of negotiations between the Division of Protection and Anthropic, which clashed over how the company might use the corporate’s fashions.
The DOD needed Anthropic to grant the Pentagon unfettered entry to its fashions for all lawful functions, whereas Anthropic needed assurance that its know-how wouldn’t be used for absolutely autonomous weapons or home mass surveillance.
The talks stalled, and the DOD declared Anthropic a provide chain danger in early March, which means that use of the corporate’s know-how purportedly threatens U.S. nationwide safety. The label requires protection contractors to certify that they do not use Anthropic’s Claude AI fashions of their work with the army.
Trump then ordered all federal companies to “IMMEDIATELY CEASE all use of Anthropic’s know-how” in a Reality Social put up, however that directive has been briefly blocked by a federal decide in San Francisco.
The DOD has not commented on Mythos, however the company has continued to make use of Anthropic’s Claude fashions within the conflict with Iran. On the very least, the facility of Mythos appears to be altering the broader tone from the Trump administration.
Bessent and Federal Reserve Chair Jerome Powell met with the heads of the highest U.S. banks final week in regards to the AI mannequin. The week earlier than, Vice President JD Vance and Bessent additionally met with Amodei, OpenAI’s Sam Altman and different high tech execs about AI cybersecurity.
Amoedi’s assembly with Wiles is the newest sign that tensions might be easing.
Wiles is a former worker of Ballard Companions, a lobbying agency with robust ties to the Trump administration. Anthropic employed Ballard after the Pentagon designated the corporate a provide chain danger, in line with federal disclosures.
— CNBC’s MacKenzie Sigalos and Megan Cassella contributed to this report.
