The United States Cybersecurity and Infrastructure Security Agency (CISA) has reportedly begun using Anthropic’s advanced artificial intelligence model, Mythos, to identify security flaws in government software, marking another major step in the growing adoption of AI for national cybersecurity.
According to reports, CISA’s Attack Surface Evaluation Team is using the AI model to scan government code repositories for vulnerabilities that could potentially be exploited by foreign hackers, cybercriminals, or espionage groups. The initiative aims to strengthen the security of critical government systems by detecting software weaknesses much faster than traditional manual audits.
Sources familiar with the project revealed that Mythos has already identified a significant number of vulnerabilities during testing, although officials have not disclosed the scale of the review or the severity of the discovered flaws. Neither CISA nor Anthropic has officially commented on the reported findings.
The development comes after a turbulent relationship between Anthropic and the U.S. government earlier this year. In February, the AI company reportedly refused to remove safeguards that prevented its AI models from being used for autonomous weapons or domestic surveillance, leading to temporary restrictions from the Pentagon. Those restrictions were later challenged in court, and relations have since improved following the private deployment of Mythos for cybersecurity applications.
Mythos has gained attention within U.S. intelligence and cybersecurity circles due to its advanced ability to identify, analyze and even simulate software vulnerabilities. Reports suggest that the National Security Agency (NSA) has also been evaluating the AI model in classified environments, with officials reportedly impressed by its performance in cybersecurity tasks.
The public version of Mythos, known as Fable, also made headlines after temporary restrictions limited access for international users due to national security concerns. Those restrictions were lifted recently, allowing wider availability while maintaining enhanced cybersecurity safeguards.
As cyber threats continue to evolve globally, governments are increasingly turning to artificial intelligence to defend critical infrastructure. AI-powered code auditing has the potential to significantly reduce the time required to detect vulnerabilities, helping security teams identify weaknesses before they can be exploited by malicious actors.
The reported use of Mythos highlights how artificial intelligence is rapidly becoming an essential component of modern cybersecurity strategies, particularly for protecting sensitive government systems against increasingly sophisticated cyberattacks.
Disclaimer: This report has been editorially prepared using publicly available information and media reports. While every effort has been made to ensure accuracy, some details may change as official statements become available. Readers are encouraged to verify important information through official sources.
Author
About the Author
