"Given the rate of AI progress, it will not be long before such capabilities proliferate, potentially beyond actors who are committed to deploying them safely. The fallout - for economies, public safety, and national security - could be severe. Project Glasswing is an urgent attempt to put these capabilities to work for defensive purposes."
"Models like Mythos will be used by 'good guys' and 'bad guys' alike. The BeyondTrust Security Team has already observed AI-assisted tooling compress the exploitation window for critical vulnerabilities to minutes, not weeks."
Claude Mythos Preview is a language model designed to identify cyber exploits, achieving high scores on vulnerability benchmarks. It has uncovered thousands of severe vulnerabilities across major operating systems and web browsers. Due to the potential misuse by cybercriminals, Anthropic has chosen not to release the model publicly. Instead, Project Glasswing aims to utilize this technology for securing critical infrastructure with select partners. Experts note that AI-assisted tools are already compressing the time to exploit vulnerabilities, highlighting the urgency of defensive measures.
Read at Securitymagazine
Unable to calculate read time
Collection
[
|
...
]