Anthropic engineers helping US deploy Mythos AI for cyber operations
What's the story
The United States National Security Agency (NSA) is using Anthropic's advanced artificial intelligence (AI) model, Claude Mythos, to plan global cyber operations, according to Financial Times. The move comes despite Anthropic's previous fallout with the Pentagon over unrestricted AI use. The tech company has also sent several Forward Deployed Engineers (FDEs) to assist in integrating Mythos into NSA systems.
AI prowess
Mythos to aid in potential cyberattacks
Mythos is considered one of the most advanced AI models in cybersecurity, capable of detecting thousands of critical vulnerabilities in widely used software. Anthropic has kept this model under wraps, citing its potential for cyberattack planning as a reason. The NSA is said to be considering using Mythos for cyberattacks against countries like China or Iran, with the aim of exploiting system vulnerabilities and gaining sensitive information.
Engineer involvement
Anthropic FDEs assisting in deployment and customization of Mythos
Anthropic has sent around half a dozen FDEs to the NSA, who are responsible for deploying and customizing AI models according to client needs. These engineers are likely helping in the deployment and customization of Mythos for specific use cases within the NSA, including potential cyberattack plans. However, it remains unclear if these engineers are directly involved in how Mythos is being used inside the agency.
AI concerns
Controversies surrounding NSA's use of Mythos
The use of Mythos by the NSA has drawn mixed reactions on social media. Some users have criticized the move as governments using AI to exploit vulnerabilities, not patch them. Others have called Anthropic hypocritical for working with the NSA after its fallout with the US Pentagon over unrestricted AI use. Despite these controversies, Anthropic plans to expand access to Mythos to about 150 more organizations in 15 additional countries, including India.