Anthropic Assists NSA with Cyber Operations While Warning on AI Risks

Anthropic has reportedly embedded about six engineers at the National Security Agency (NSA) to deploy its Mythos AI model for offensive cyber operations. These operations might include targeting networks in countries like China and Iran, according to a report by the Financial Times. The embedding of these engineers aims to customize Mythos for specific applications, though it remains unclear whether they are involved in active military operations.

Mythos, Anthropic's most capable AI model, has not been publicly released due to concerns over potential misuse. Instead, it is available to vetted partners through Project Glasswing, which includes major technology firms such as Microsoft, Apple, and Amazon. Anthropic is currently engaged in a legal dispute with the Pentagon after a $200 million contract with the Department of Defense fell through, stemming from conflicts over usage limitations, especially concerning autonomous weapons and mass surveillance.

On the same day its collaboration with the NSA was disclosed, Anthropic's research institute published a report titled "When AI Builds Itself," which discusses the rapid development of AI capabilities like self-improvement. In conjunction with this, Anthropic has called for a coordinated global pause on AI advancements. This call to action aims to mitigate risks as AI technologies increasingly become capable of autonomous development.

These developments are particularly significant as Anthropic is also filing for an initial public offering (IPO) that could value the company at over $1 trillion. The potential for AI to enhance its own development raises critical questions about control and safety in a rapidly evolving technological landscape.

Going forward, industry stakeholders and regulators will need to monitor how these AI models are used, particularly in sensitive contexts such as national security. As Anthropic navigates its legal battles and prepares for its IPO, the implications of its technologies on both commercial and governmental fronts will be closely scrutinized.