We evaluate several models on 100 tasks from the [internal Binary Exploitation benchmark] (selected at random), and find that GLM-5.3 develops full control flow hijacks in 4% of the trials; Claude Mythos Preview did so in 6%.
Quoting Anthropic Frontier Red Team
About this summary. This is a short, independently written summary of an article first published by Simon Willison. Cyber Security News did not report or verify the underlying story. Read the original: https://simonwillison.net/2026/Sep/29/anthropic-frontier-red-team/
Source attribution: headline and facts are from Simon Willison (simonwillison.net). Summary method: excerpt of the source description. See our source attribution policy.
Related coverage
OpenAI reveals ‘novel’ encryption bypass used in distillation attack
OpenAI said it disrupted a “coordinated campaign” to distill and extract reasoning capabilities from its AI models, pointing the finger at a Chinese rival.
Irony alert: OpenAI whines that Chinese model stole its special IP that it stole from everybody else
OpenAI, which hoovered up vast amounts of internet content amid copyright fights, has accused individuals associated with China’s Moonshot AI of being involved in a…
China-nexus UAT-11587 targets government and policy organizations across Asia with Antino backdoor
Cisco Talos uncovered a cluster of activity we track as UAT-11587 targeting government and policy organizations across Asia, including in Taiwan, India, the Philippines…
WaterISAC reckons with range of threats after summer of cyberattacks
The computers that automate water treatment systems across the country were built for durability, not for an internet-connected world. Many still accomplish their…
When Privacy Becomes a Weapon: Understanding Doxxing and Privacy Vulnerabilities in Mainland China's Social Media Ecosystem
arXiv:2609.35951v1 Announce Type: new Abstract: Doxxing, the malicious disclosure of personal information, has become a pervasive privacy threat. Yet existing research…
'NeedyMantis' Provides Long-Term Access to Compromised Networks
Microsoft observed a China-based actor using a previously unidentified malware framework in targeted intrusions against telcos, universities, medical, and…