Cybersecurity vets protest ‘dangerous’ US government ban on Anthropic’s most powerful models

Dozens of cybersecurity experts sent a letter to the White House this week protesting the U.S. government's ban on Anthropic's most powerful AI models, Fable and Mythos. The experts contend that these export control restrictions will hinder the ability of cybersecurity defenders to secure software and products. They argue that restricting access to these advanced AI tools prevents researchers and security professionals from using them to identify vulnerabilities and develop better defenses against cyber threats. The letter, which was not publicly released, urges the administration to reconsider the ban, emphasizing that open access to these models is crucial for advancing cybersecurity research and protecting critical infrastructure. The experts believe that by limiting the availability of these tools, the U.S. government is inadvertently weakening its own cybersecurity posture and that of its allies. They propose that instead of a ban, the government should focus on developing robust frameworks for responsible use and monitoring of these powerful AI systems.