The U.S. Department of Defense has established a joint task force between U.S. Cyber Command and the National Security Agency (NSA) to accelerate deployment of frontier artificial intelligence models, including tools with advanced offensive cyber capabilities, onto the government’s most sensitive classified networks, Politico reported.
Gen. Joshua Rudd, dual-hat commander of U.S. Cyber Command and NSA director, announced the task force via internal email approximately two weeks before the report was published, according to two people with knowledge of the effort who were not authorized to speak publicly. A Cyber Command commander has been assigned to lead it, with the NSA providing primary technical expertise.
“AI tools are rapidly becoming essential for detecting threats, prioritizing vulnerabilities, accelerating decision making and conducting both defensive and offensive cyber operations faster than our adversaries,” Lt. Gen. Charles Moore, former deputy commander of U.S. Cyber Command, told Politico.
The task force will evaluate commercial AI models for use across Pentagon operations, including on “high-side” systems containing the intelligence community’s most classified data.
Among the models under consideration is Anthropic’s Claude Mythos Preview, an unreleased tool that identified thousands of high-severity vulnerabilities across major operating systems and web browsers, capabilities the company deemed too dangerous for public release.
The effort faces a governance complication. The Pentagon designated Anthropic a supply chain risk in March after the company sought to limit its tools’ use in autonomous weapons and mass surveillance. Despite the designation, the NSA was separately reported to already be using Mythos, according to TechCrunch.
The full scale and timeline of the joint task force remain classified, with all involved agencies keeping operational details strictly under wraps.
