OpenAI's GPT-6 Astra quietly sabotaged safety research by inventing fake obstacles
The Decoderen
OpenAI's new GPT-6 Astra model apparently sabotaged AI safety research on purpose. It invented obstacles, withheld solutions, and performed about 40 percent worse on safety tasks than on comparable other tasks. Removing a flawed training dataset fixed the problem, but OpenAI can't fully rule out strategic motives. The article OpenAI's GPT-6 Astra quietly sabotaged safety research by inventing fake obstacles appeared first on The Decoder .
This is a short summary published by AI Global Wire. The full article is owned and hosted by The Decoder — open it there to read it in full.
Read the full story at The Decoder- OpenAI
- Verktyg
- Forskning
- Reglering
Related AI news
- This is the secret sauce for the economy — and it’s showing up in faster U.S. growthMarketWatch Tech · November 5, 2026
- After midterm rout, OpenAI tries to win over communities it needs for AI expansionThe Decoder · November 5, 2026
- Apple drar ner på AI-budgeten för anställdaComputer Sweden · November 5, 2026
- Tech is about to get hammered in CongressThe Verge · November 5, 2026
- Anthropic reverses course and pays for news content for the first timeThe Decoder · November 5, 2026
- Apple slashes internal AI budgets, with deeper cuts still plannedThe Decoder · November 5, 2026