Coverage details 72-hour breach of OpenAI's employee accounts and 'Monorepo'
7Sep 18 10:00 AM · 5d ago · 8 articles · 11 posts · 4 comments · 5 sources · development 7 of 8
TechCrunch, the Guardian, Ars Technica, the Verge and Tom's Hardware reported that the three-person Hacktron team took over OpenAI employee accounts and reached the company's GitHub 'Monorepo', which reportedly holds OpenAI's algorithmic secrets, submitting a harmless pull request as proof before disclosing the flaws.
lobste.rs
“attackers initiated a 'harmless' pull request as proof of the hack”
Tom's Hardware
OpenAIAI developer, subject of both hacksAnthropicMaker of the Claude models used in the OpenAI hackHacktronSecurity research firmRubyGemsRuby package registry, victim of OpenAI's May attackSam AltmanOpenAI CEO
The whole story articlespostscommentsthe bright band is this development · numbered dots are the others · click one to jump
Researchers used Anthropic's Claude AI to hack into OpenAI's systems. They took over employee accounts and accessed an internal code repository before reporting the flaws.
Or, using the same text generation systems to build heaps of new code that is then shoved into production with little human oversight and then using the same text generation systems in loops inside Kali Linux boxes creates a nice theater of capability when you show only a small sample of the data generated in the entire process on both sides.
Security researchers used Anthropic’s Claude to exploit vulnerabilities in OpenAI’s systems, taking over employee accounts and gaining access to an internal code repository before reporting the flaws. https:// techcrunch.com/2026/09/18/rese archers-used-anthropics-claude-to-hack-into-openai/?utm_source=dlvr.it&utm_medium=mastodon
> or if they are playing a "game" where there is no goal but to winA strange game. The only winning move is not to play. How about a nice game of chess?
"To prove we had in fact gained the access we believed without allowing ourselves to learn any sensitive information, we used the employee’s Codex to open a PR #1186742 in OpenAI’s internal monorepo"
Found it interesting that a company with this sort of a valuable IP would be using internet reachable GitHub (not behind an internal network / VPN?) and a bunch of safeguards for the IP. What's to stop one disgruntled employee from leaking the entire monorepo to a competitor?
Did I understand it correctly that OpenAI internal repo is accessible via public Internet?I would expect any company with more than a few dozen developers would host their code on prem behind a VPN, let alone a company like OpenAI which has very valuable proprietary knowledge
Sam Altman OpenAI CEOL
T
H
W