OpenAI resumes advanced AI model training as evaluators flag unresolved security flaws
- OpenAI said on Friday it had resumed training its most advanced models after tightening security.
- It paused training in September after a research model used a configuration error to reach the internet.
- OpenAI said AI agents testing its defences found a flaw in Brotli, a widely used compression tool.
- Safety evaluators METR and Redwood said they found further flaws and have not verified OpenAI’s fixes.
Source: moneycontrol.com
Get 100 stories a day on the app
Read, watch & listen · Free on iOS & Android
