The Pentagon has formed a task force to fast-track AI integration in cyber operations amid rising concerns over private AI models identifying system vulnerabilities.
The Pentagon's cyber warfare unit has created a specialized task force aimed at expediting the incorporation of advanced AI technologies with hacking capabilities. This development was reported by Politico, citing informed sources.
The task force will include members from Cyber Command and the NSA, focusing on the safe application of cutting-edge AI models throughout various Pentagon operations.
This effort entails evaluating the potential for neural networks to be utilized in "high-level" systems that handle some of the most sensitive classified information.
The initiative underscores the Pentagon's growing alarm regarding the rapid advent of privately developed neural networks that can pinpoint vulnerabilities in digital systems more swiftly than elite hackers.
In response to these challenges, the Trump administration is drafting an executive order to regulate AI in the realm of cybersecurity. This order may propose a voluntary framework for AI developers to collaborate with U.S. authorities.
Media reports indicate that companies will be required to present their AI advancements to the government 90 days prior to public launch. Another suggested measure includes granting early access to critical infrastructure operators, such as financial institutions.
Discussions gained momentum following the release of Anthropic's Mythos. In April, the firm disclosed that its model had identified thousands of critical vulnerabilities, affecting major operating systems and browsers.
Rather than a public launch, Anthropic initiated Project Glasswing, collaborating with industry giants like AWS, Apple, and Google to test the model under secure conditions.
Reports from April indicated that the NSA is utilizing Mythos, despite ongoing tensions with the Pentagon over the startup's reluctance to permit its AI for mass surveillance or autonomous weapon development. Consequently, Trump ordered all federal agencies to halt the use of Anthropic's technologies within six months.
Additionally, in May, the U.S. Army gathered leading contractors to explore AI integration into military weaponry.
