President Donald Trump has ordered federal agencies to immediately halt the use of technology developed by Anthropic, escalating a standoff between the White House and the San Francisco–based artificial intelligence company.
The directive followed Anthropic’s refusal to agree to the Pentagon’s request for unrestricted military access to its Claude AI models. The company said it would not permit its systems to be used for mass surveillance of Americans or for fully autonomous weapons. In a strongly worded statement, Anthropic insisted that pressure from the Defense Department would not force it to abandon those limits.
Trump responded on his Truth Social platform, declaring that every federal agency must cease using Anthropic’s technology at once. He said the government no longer needed the company’s products and warned of serious civil and criminal consequences if Anthropic failed to cooperate during the transition. He also suggested he would use the full authority of the presidency to enforce compliance if necessary.
The Pentagon had earlier set a deadline for the company to accept its terms, warning that failure to do so could trigger action under the Defense Production Act. That Cold War-era law, which was also invoked during the Covid pandemic, gives the federal government broad powers to direct private companies in matters tied to national security. Defense officials also raised the possibility of labeling Anthropic a supply chain risk, a designation usually applied to firms linked to hostile foreign governments.
Anthropic signaled it would challenge any such move in court. The company argued that while artificial intelligence can support national security in some cases, it can also threaten democratic values if misused. It maintained that clear boundaries are necessary, particularly when it comes to domestic surveillance and autonomous weapons.
US Defense Secretary Pete Hegseth said he had instructed the Pentagon to move forward with restrictions, adding that contractors and suppliers working with the US military would not be permitted to engage in commercial activity with Anthropic. He sharply criticized the company’s stance, accusing it of undermining its relationship with the government.
The dispute has drawn political and industry-wide reactions. House Democratic leader Hakeem Jeffries defended Anthropic’s position, describing the Pentagon’s approach as an invasion of privacy and warning that mass surveillance of American citizens is unacceptable.
Within the tech sector, the confrontation has sparked solidarity. Hundreds of employees from Google DeepMind and OpenAI signed an open letter urging their employers to stand together and resist pressure to deploy AI systems for domestic surveillance or autonomous lethal operations. The letter argued that dividing companies through fear would only weaken collective efforts to set ethical limits.
OpenAI’s chief executive, Sam Altman, told staff he was seeking an arrangement with the Pentagon that would include clear red lines similar to those proposed by Anthropic. He emphasized the need for human oversight in high-stakes decisions and reiterated opposition to the use of AI for mass surveillance or fully autonomous weapons.
Anthropic, while firm in its position, said it remains willing to work with the US government on projects that strengthen national security within ethical boundaries. The standoff now moves toward what could become a significant legal and political battle over how far artificial intelligence should go in military and domestic use.
