NEW YORK/WASHINGTON/SAN FRANCISCO: Even after U.S. Defense Secretary Pete Hegseth issued an ultimatum to Anthropic to get on board and comply with U.S. law, the artificial intelligence lab said it had no intention of easing its usage restrictions for military purposes, a person familiar with the matter said on February 24.

The meeting between Anthropic CEO Dario Amodei and U.S. Defense Secretary Pete Hegseth was scheduled to hash out a months-long dispute. However, people familiar with the matter said Hegseth delivered an ultimatum to Anthropic during the meeting: Get on board or the government would take drastic action.

Hegseth warned that the Pentagon could use the Defense Production Act to force Anthropic to meet its demands, or label the company as a supply chain risk — a label usually given to companies linked to hostile foreign countries. This could hurt Anthropic's business with other companies that work with the U.S. government.

The AI startup has refused to remove safeguards that prevent its technology from being used to target weapons autonomously or to conduct U.S. domestic surveillance.

A senior Pentagon official said Anthropic had until 5:00 p.m. on February 27 to respond to the government's ultimatum.

The Pentagon did not respond right away to a request for comment. An Anthropic spokesperson said that the February 24 meeting was part of ongoing, honest discussions about its usage policy, so the company can continue helping with national security safely and responsibly.

The Pentagon has been discussing AI contracts with several major large language model (LLM) companies, including Google, xAI, and OpenAI. These deals will help determine how the military uses artificial intelligence in the future, including for drone swarms, robots, and cyberattacks.

Until recently, Anthropic was the only LLM company working on classified government networks. This week, the Pentagon said it had made a deal with xAI to use its technology on classified networks as well. Reuters earlier reported that the Pentagon plans to allow all AI companies onto these secure networks.

Tensions between the Pentagon and Anthropic increased earlier this month. The Pentagon became worried that Anthropic had asked questions about how its AI tools were used during a military raid in Venezuela that led to the capture of President Nicolas Maduro.

At the meeting with Hegseth, Anthropic CEO Dario Amodei said the company had not raised concerns with Palantir or the Pentagon about whether its AI was used in the Venezuela raid, according to a source. He also said that the company's current safety measures would not interfere with the Defense Department's operations.

A government contracts lawyer, Franklin Turner, said this situation is highly unusual and would likely lead to many lawsuits if the administration takes action against Anthropic.