"Defenders finally have a chance to win, decisively. Elite security researchers find bugs that fuzzers can't largely by reasoning through the source code."
"Computers were completely incapable of doing this a few months ago, and now they excel at it. Mythos Preview is every bit as capable as the world's best security researchers."
"For a hardened target, just one such bug would have been red alert in 2025, and so many at once makes you stop to wonder."
Mozilla's Firefox 150 release addresses 271 security vulnerabilities discovered through Anthropic's Claude Mythos Preview AI model. This model, not publicly available, was accessed by Mozilla through a collaboration with Anthropic. Unlike traditional fuzzers, Mythos identifies bugs without blind spots by reasoning through source code. Mozilla's Chief Technology Officer, Bobby Holley, emphasized the model's capability, stating it matches elite security researchers. The initial findings at Mozilla were overwhelming, raising concerns about the number of vulnerabilities found in a single release.
Read at TechRepublic
Unable to calculate read time
Collection
[
|
...
]