U.S. Agency Deploys AI Model to Enhance Cybersecurity Protocols
WASHINGTON, D.C.: The Cybersecurity and Infrastructure Security Agency (CISA) is leveraging Anthropic’s artificial intelligence framework, Mythos, to scrutinize government software for potential security vulnerabilities, according to three informed individuals.
This initiative underscores an increasing governmental inclination towards integrating innovative AI solutions.
The agency is utilizing Mythos to meticulously analyze government code repositories, seeking out weaknesses that may be exploited by foreign intelligence operatives or cybercriminal entities, the sources disclosed.
While Anthropic has not addressed inquiries regarding this endeavor, a CISA spokesperson indicated last month that the agency would verify any shareable information, yet did not respond to follow-up requests for elucidation.
Per one source, the operations are orchestrated by CISA’s Attack Surface Evaluation team, which undertakes comprehensive digital security assessments and hacking simulations across various sectors of the U.S. government.
Two sources mentioned that the audits have already identified a substantial array of vulnerabilities, although specific details remain undisclosed.
It remains unclear the extent of government code examined or the intricacies and seriousness of the identified flaws.
This initiative surfaces amid Anthropic’s complex rapport with the U.S. government.
The company, which confidentially filed for an initial public offering in the United States, faced pressure earlier this year after opting not to dismantle safeguards that barred its AI systems from applications in autonomous weaponry or domestic surveillance.
This contention prompted the Pentagon, in February, to categorize Anthropic as a supply-chain risk—a designation typically reserved for foreign entities suspected of engaging in espionage.
A judicial intervention halted this classification in March, and subsequent tensions have subsided since the private unveiling of Mythos, an AI model aimed at detecting and exploiting cybersecurity vulnerabilities.
The National Security Agency has also been employing Mythos since at least April, despite the prior blacklisting, as reported by Axios.
Additionally, a recent New York Times article indicated that NSA analysts tested the model within classified settings and expressed admiration for its adeptness.
Upon the public release of a variant of Mythos, dubbed Fable, which incorporated cybersecurity safeguards, the White House mandated that foreign users be denied access.
This directive led to a brief global suspension of the model, which was subsequently relaxed last week. The NSA and the White House have not yet provided a response to requests for comment.
Source link: Indiasnews.net.
