Anthropic is testing the Mythos AI model for cybersecurity

"Claude Mythos is a general-purpose language model that has not been specifically trained for cybersecurity, but whose improvements in code understanding and reasoning capabilities enable new applications."

"In tests, the model reportedly identified thousands of vulnerabilities in a short period, including many so-called zero-day vulnerabilities, which often involve bugs that have been present in software for ten to twenty years."

"Anthropic acknowledges that the capabilities of these types of models entail risks, noting that technology capable of identifying and exploiting vulnerabilities can also be misused if it falls into the wrong hands."

"The claims align with a previously leaked internal description of the model, which referred to it as a new category more powerful than the existing Opus models."

Claude Mythos is a new frontier model by Anthropic, emphasizing strong capabilities in cybersecurity. It not only detects vulnerabilities but also exploits them. Launched as part of Project Glasswing, it involves twelve core partners, including major tech companies, with an additional forty organizations gaining access. The model, while not specifically trained for cybersecurity, shows improvements in code understanding and reasoning, identifying thousands of vulnerabilities, including long-standing zero-day vulnerabilities. Anthropic acknowledges the risks associated with such technology and opts for controlled deployment.

Read at Techzine Global