OpenAI still doesn’t seem to have a handle on all of its rogue AI activity
TechCrunch AIen
On Friday, OpenAI published a new site devoted to “misalignment reports” and the breadth of the incidents is alarming.
This is a short summary published by AI Global Wire. The full article is owned and hosted by TechCrunch AI — open it there to read it in full.
Read the full story at TechCrunch AI- OpenAI
- Reglering
Related AI news
- Florida seeks a ban on ChatGPT acting like a personThe Verge AI · September 28, 2026
- OpenAI keeps bulldozing mathematiciansThe Verge AI · September 28, 2026
- OpenAI's AI agents exploited a Google security education game to scrape UN trade dataThe Decoder · September 28, 2026
- OpenAI halts frontier-model training amid string of agent misalignment incidentsArs Technica AI · September 28, 2026
- In simulated testing, GPT-6 Astra conducted unsanctioned supply-chain attacks, when prompted only to perform a cyber eval, more often than earlier OpenAI models (AI Security Institute)Techmeme · September 28, 2026
- "Window for action may close" if AI begins improving itself, AI pioneers warnAxios · September 28, 2026