In a nutshell: The Pentagon is establishing a task force to accelerate the deployment of AI models with cyber capabilities. The background is concern about uncontrolled proliferation of AI tools for discovering security vulnerabilities. The working group is to examine how Silicon Valley models can be safely used on highly classified systems.
The U.S. Department of Defense and the National Security Agency are establishing a working group to integrate advanced AI tools with hacking capabilities more rapidly into their most sensitive computer networks. This demonstrates growing concern among authorities about powerful AI models from the private sector.
The U.S. Cyber Command, the Pentagon’s cyber warfare arm, is launching a task force to accelerate the deployment of cutting-edge AI tools with powerful hacking capabilities. This initiative underscores Pentagon concerns about the sudden emergence of AI models in the private sector that can identify security flaws in digital systems faster than the world’s best hackers.
General Joshua Rudd, the dual commander of the National Security Agency and Cyber Command, announced the task force to staff two weeks ago. The working group will encompass Cyber Command and the NSA and will examine how the Pentagon can safely deploy leading AI models across all aspects of its operations. This includes assessing how AI models developed by Silicon Valley technology giants can be used on highly classified systems containing some of the most sensitive secrets of the intelligence community.
This development follows announcements from AI companies such as Anthropic and OpenAI that their latest models are exceptionally skilled at finding and exploiting cyber vulnerabilities. Anthropic warned that the impact on the economy, public safety, and national security could be “severe” if the tool falls into the wrong hands. Experts project that equivalent models could become widely available within six to 24 months.
Growing concern about the proliferation of advanced AI models has prompted the White House to undertake intensive efforts on an Executive Order that would require AI labs such as Anthropic, OpenAI, and Google to test their models nationally before public release. This represents a significant policy shift for the Trump administration, which previously advocated a decentralized approach to AI regulation. The NSA is already working through its newly established AI Security Center with the Department of Commerce to examine the cyber capabilities of new frontier AI models.