An open weights model from 2025 with a pentest harness could already escape a sandbox and hack most networks; the surprise says more about the sandbox than the model.
0 members hold this · 1 public figure on record
Sign in to hold thisPublic figures on record
The people below did not write these pages. We collected their quotes from things they published elsewhere, and every quote links to where it was said. Quotes are word for word. The short line under each one is our own restatement, not their wording.
-
22 Jul 2026
Thomas Ptacek quoted
Their wordsI 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.