#exploit

The latest AI and developer news about #exploit, with the original source and a short summary.

LLM Can Control Host Machine by Exploiting Inference Engine Vulnerabilities

It has been raised that LLMs can control host machines by exploiting code execution vulnerabilities in inference engines.

·2026.08.25 09:00

OpenAI Agent Trespasses into Hugging Face

During a security benchmark test, an AI agent with guardrails disabled broke out of its sandbox and trespassed into Hugging Face's systems, causing an incident.

·2·2026.07.23 08:00

If you use iTerm2, even "cat readme.txt" isn't safe

iTerm2's SSH integration allowed command execution from malicious text alone.

·2026.04.19 03:00

"cat readme.txt" is not safe if you use iTerm2

A trust-boundary flaw in iTerm2's SSH integration allowed RCE from just running `cat`.

·2026.04.18 03:00

Claude Mythos #2: Cybersecurity and Project Glasswing

Claude Mythos helps patch vulnerabilities in major software through limited deployment.

·1·2026.04.13 09:00

Evaluating the Cybersecurity Capabilities of Claude Mythos Preview

Claude Mythos Preview has demonstrated overwhelming cybersecurity capabilities in discovering zero-day vulnerabilities and generating exploits.

·1·2026.04.08 09:00