Anthropic's Mythos Preview was highly effective at finding vulnerability candidates, especially when analyzing source code.
CISA added CVE-2026-42271, a high-severity LiteLLM command injection flaw, to its KEV catalog after evidence of active ...
Hackers compromised 19 packages on the PyPI, collectively downloaded hundreds of thousands of times, in a new Shai-Hulud ...
A flaw in Hugging Face Transformers could allow malicious AI models to execute code, exposing credentials and highlighting AI ...
Public LLM models with safeguards turned off can also build working exploits, increasing patch gap risks. Anthropic says its Claude Mythos Preview model can build working exploits targeting known ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results