Anthropic flags cybersecurity risks ahead of IPO

In its SEC filing ahead of an IPO, Anthropic warned that its Claude models could be misused for cyberattacks by foreign governments and criminals, exposing it to lawsuits and insurance problems. It also cited export controls on its Mythos 5 and Fable 5 models and an FTC civil investigative demand over product safety and privacy.

• Anthropic cited liability and insurance risks from misuse of its models
• US export controls covered the Mythos 5 and Fable 5 models
• The FTC requested information on product safety and privacy policies
• AI agents create new attack surfaces via third-party APIs and databases