Anthropic clashes with Pentagon over AI guardrails as $200 million contract hangs in the balance
Leading artificial intelligence research company Anthropic is facing mounting pressure from the U.S. Department of Defense to remove safety restrictions from its AI system, Claude
(CNN)- Leading artificial intelligence research company Anthropic is facing mounting pressure from the U.S. Department of Defense to remove safety restrictions from its AI system, Claude — the first AI tool approved for use on the military’s classified network.
On Tuesday, Pete Hegseth told Anthropic CEO Dario Amodei that if the company does not allow Claude to be used for “all lawful purposes,” the Pentagon would cancel its $200 million contract and label the firm a “supply chain risk.”
But in a blog post Thursday, Amodei rejected the demand, warning that artificial intelligence could “undermine, rather than defend, democratic values” if deployed without proper safeguards. He said Anthropic would not change its position despite the threat.
Tech journalist Kara Swisher said concerns center around how the technology could be used — including for surveillance and autonomous weapons systems.
“They’re worried about how this technology is deployed,” Swisher said. “These are not perfect systems. What if it did do something wrong?”
Some analysts say Anthropic could strengthen its brand by standing firm, noting the company has built its reputation around AI safety and responsible deployment.
The battle over artificial intelligence this week extends beyond Washington.
Block, Inc. — the parent company of Square, Inc., Cash App and Afterpay — announced thousands of layoffs, cutting its workforce by 40 percent.
In a letter to shareholders, co-founder Jack Dorsey cited the rise of “intelligence tools” as a key reason for the reductions.
CNN tech reporter Claire Duffy said the company believes artificial intelligence can perform some of the work more efficiently than human employees — a move experts warn could signal broader workforce shifts across multiple industries.
As the debate over AI safety, national security and job displacement intensifies, the outcome of Anthropic’s standoff with the Pentagon could set a precedent for how far the government can push private tech firms in the rapidly evolving AI arms race.
