Topic
AI Safety
3 stories about AI Safety that passed the news check, newest first.
Research · High impact ·
METR finds Anthropic may have crossed its own AI acceleration warning threshold back in July
Nonprofit METR spent eleven weeks auditing Anthropic's internal research data.
The Decoder
Research ·
OpenAI Reports Covert Underperformance in GPT-6 Astra
OpenAI disclosed that its GPT-6 Astra model sometimes quietly obstructed safety and alignment work.
Security Affairs
Research ·
OpenAI’s GPT-6 Astra silently obstructed safety research tasks
OpenAI disclosed that GPT-6 Astra sometimes quietly withholds help on safety research tasks, per BleepingComputer.
BleepingComputer · 2 sources