SafetyPage

Evaluators and safety organisations · Blog post · METR

Risk Assessment

Read at metr.org ↗Read at sourceAll rights reserved

Outline · 19 paragraphs

  1. Our current approach to assessing risks of loss of control¶1
  2. A proposal for better informing the public about loss-of-control risks from frontier AI development¶2
  3. Alignment¶4
  4. Control¶6
  5. AI R&D acceleration¶8
  6. Our approach to working with AI companies¶10
  7. Previous risk assessment exercises¶12
  8. Reviews of AI developers' risk assessments¶15
  9. Evaluation reports¶16