Anthropic said it blocked its most advanced artificial intelligence models for public use after the U.S. government banned their use by foreigners.
Anthropic released its Fable 5 and Mythos 5 models earlier this week, advertising them as the most advanced AI models given to the public to date. On Friday, the company received a directive from the U.S. government banning the models’ use by any foreign national, citing concerns over an alleged method to “jailbreak,” or bypass, the models’ security features. To comply with the demand, Anthropic suspended use of the model for all users, claiming that was the only way to cooperate with the order.
Recommended Stories
“We are complying with the government’s legal directive and are removing access to Fable 5 and Mythos 5 for all users,” a statement from the company said. “However, we disagree that the finding of a narrow potential jailbreak should be cause for recalling a commercial model deployed to hundreds of millions of people. If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers.”
Anthropic added that access to all other Claude models shouldn’t be affected by the move.
The company questioned the government’s logic, downplaying the severity of the “jailbreak.”
AI GIANT ANTHROPIC FILES TO GO PUBLIC
“To date, the government has only given us verbal evidence of a potential narrow, non-universal jailbreak, which essentially consists of asking the model to read a specific codebase and fix any software flaws,” the statement said.
The episode marks the latest feud between the U.S. government and Anthropic. They have clashed repeatedly over the government’s authority over AI companies, particularly in the military realm. Earlier this year, Anthropic had a high-profile standoff with the Pentagon over restrictions on the use of its models.
