"360 Digital Security Group's internally developed 'Multi-Agent Collaborative Vulnerability Discovery System' played a significant role in its first-place finish at the Tianfu Cup, identifying close to 1,000 vulnerabilities, including over 50 high-severity flaws across various platforms."
"The most notable claim involves CVE-2026-32190, a critical Office vulnerability that 360 asserts its AI agent identified within minutes, after allegedly going undetected for roughly eight years."
A Chinese cybersecurity firm, 360 Digital Security Group, asserts that its AI-driven system can discover vulnerabilities at a level comparable to Anthropic's Claude Mythos model. This claim follows Anthropic's announcement of its Mythos model, which autonomously identified thousands of vulnerabilities but is restricted to select organizations. The 360 system reportedly contributed to half of the vulnerabilities identified in a major hacking competition, including a critical Office vulnerability that had remained undetected for eight years.
Read at SecurityWeek
Unable to calculate read time
Collection
[
|
...
]