OpenAI built a model so good at hacking, it cannot release it.
This is a dev post classified by AGK Intelligence as Security (a launch), kept by the Dev Radar because it carries real work, not commentary.
🔥🔥 OpenAI built a model so good at hacking, it cannot release it. GPT-6 Cyber found 81 previously unknown high-severity bugs in Firefox, and helped build working exploits for 948 Linux kernel vulnerabilities. It is also much stronger than GPT-5.6 Cyber, and it stayed inside the rules while doing it. That is the fix for the Astra problem, where the model got better at breaking things and worse at staying in bounds. That is why it is rated Critical, OpenAI's highest cyber tier, and why it is only going to vetted Daybreak Red customers. The best hacker OpenAI has built is locked in a room.
Posted by Vaibhav Sisinty (177.7k followers) 1 h ago · 4 likes · 1.2k views · view the original post on X. Kept by the Dev Radar as Security. Tools mentioned: openai.com.
More dev work like this
- BREAKING: OpenAI just named the model behind yesterday's Linux kernel run. — @guilleflorvs
- Cybersecurity experts right now — @karankendre
- wanna see how AI is changing the speed of research? — @haider1
- doing serious numbers — @teortaxesTex
- We put GPT-6 Cyber inside CodeRabbit’s security scan pipeline. — @coderabbitai
- GPT 6 CYBER: not available to peasants — @Hesamation
- GPT-6 Cyber: a major advance for security research, already being used to help defend… — @gdb
- Strongest Cyber model yet from OpenAI, it outperforms Astra. Also, OAI are creating a… — @AndrewCurran_
Every post is read and classified by AGK Intelligence: what it is, which market it belongs to, and whether the link is a real tool. 31.6k posts from 5k X accounts over the last 21 days, 9.4k tools, 12 markets. New posts are collected several times an hour and the full ranking is recomputed every hour — last update 2026-11-05 18:57 UTC. Full method.