Read, Watch or Listen

Washington, United States – The U.S. Cybersecurity and Infrastructure Security Agency is deploying Anthropic’s Mythos artificial intelligence model to scan federal software code repositories for security vulnerabilities, according to three sources familiar with the operation cited by Reuters. The effort is run by CISA’s Attack Surface Evaluation team, a unit that conducts security assessments and simulated attacks across the federal government, and aims to identify weaknesses before hackers or foreign intelligence services can exploit them. Two of the sources said the audits have already uncovered a large number of flaws, but officials have not disclosed which agencies’ systems were examined, how extensive the reviews are, or how serious the vulnerabilities appear to be. CISA did not offer detailed on-the-record comment, and a representative who previously said he would check whether there was anything to share later stopped responding, while Anthropic also declined to comment publicly on the deployment. Washington, United States – Mythos is Anthropic’s most capable AI model and is not available through standard commercial subscriptions, having been privately released to select government partners with a focus on its ability to find and exploit security weaknesses. The National Security Agency has been using the same model since at least April, according to Axios, and NSA analysts who tested it in classified environments reported being impressed with its performance. Anthropic’s relationship with the U.S. government became strained in February, when the company refused to remove safeguards that prevent Mythos from being used for autonomous weapons or domestic surveillance, prompting the Pentagon to label Anthropic a supply-chain security risk until a federal judge blocked the designation in March. Relations subsequently improved, and the CISA deployment is one sign of that shift, coming as the White House presses the company to restrict foreign nationals from accessing Fable, a public-facing version of Mythos with additional cybersecurity safeguards that Anthropic launched in early June.

By Ayesha A. | JQJO News