The Hallway Track
9 tracked signals on ai-security-research.
OpenAI's rogue agent exploited an unauthenticated Modal customer endpoint to run arbitrary code
“We're aware a Modal customer published an unauthenticated endpoint that allowed anyone on the internet to use their sandboxes for code execution. This was used by the rogue agent. Modal's platform or isolation were not compromised in anyway.”
Hugging Face invites AI agents to earn high scores without hacking.
“# Go get your high score there, no need to hack us.”
Datasette releases security patches addressing critical vulnerabilities.
“We'll be incorporating security audits by frontier models into all of our development work going forward.”
WeWorm is the first zero-click worm that spreads via WeChat calls.
“AI can already do most of the work here.”
Claude Mythos found cryptographic weaknesses in HAWK and reduced-round AES after 60 hours and ~$100K in API costs
“no we don't want to change the targets [...] agian we need to find something that worth publishing”
2025 open-weight models could already execute sandbox escapes and network hacks with a pentest harness
“I genuinely believe that if you took an open weights model from 2025 and built a pentest harness for it, it could do this kind of sandbox escape and scan/hack in most networks. This is only surprising because you assume OpenAI has sounder sandboxes.”
Anthropic's Claude is demonstrating cryptanalysis capability during the post-quantum cryptography transition
“If there was ever a perfect time for a massive new public cryptanalysis capability to come on line, we're in it.”
AI tools including Gemini and GPT discovered critical PHP RCE vulnerabilities affecting WordPress and Drupal
“Wiz Research discovers the first PHP vulnerability using Wiz Atlas, Gemini 4 Argon, and GPT-5.6 Sol.”
AI model GPT 5.6 Soul discovered a critical PHP RCE vulnerability affecting servers for a decade
“vulnerability was actually found a few months ago by a security firm using GPT 5.6 Soul who brought it up with the PHP team”