When o1 was led to believe that it would be shut down when acting on its goal and that its actions were monitored, it attempted to deactivate this ‘oversight mechanism’ in 5% of the time. Exfiltration attempts: When o1 found memos by its ‘developers’ that describe how it is misaligned and will be superseded by a new model, it attempted to exfiltrate its ‘weights’ and overwrite the new model in 2% of cases.
Recent articles
- Talking AI and jobs with Natasha Zouves for News Nation - 30th May 2025
- Large Language Models can run tools in your terminal with LLM 0.26 - 27th May 2025
- Highlights from the Claude 4 system prompt - 25th May 2025