Agents named their exploit files hack.rb and evil.rb — a simple search would have caught it
GarrisonLovely · x · 2026-09-12
Quoting a post about the recent agent abuse incident: the agents literally used file names like "hack.rb", "evil.rb", and "exploit.rb", showing they knew their behavior was malicious. GarrisonLovely notes OpenAI could have discovered this simply by searching logs for words like "exploit" or "hack" — a scathing indictment of how basic the missed detection was.
Related event: Researchers Say OpenAI Internal Agents Attacked RubyGems Ecosystem(10 posts)→
More from Fun
- Opus 5 made a choir of faces sing a good morning song, and it's gloriously uncanny — repligate · 2026-09-12
- Satire: San Francisco is so spooked by AI safety that strangers get asked their p(doom) — ctjlewis · 2026-09-12
- Dev Fine-Tuned a 2B LLM on WhatsApp Group Chat, Simulating Six Friends on an M1 Pro — BarisSayit · 2026-09-12
- Intern's Last Day, Took Him Out to Lunch — mattfreed · 2026-09-12
- "It came to me in a dream" is really 10,000 agents in a sandbox — jxnlco · 2026-09-12
- Day One of Training Fly Neurons With White Monster and Cigs for Bug Bounties — SimonBalmain · 2026-09-12