Skip to content
K = 0.731

OpenAI Shares METR and Redwood Research Assessment of Internal Agent Monitoring

1 reports 1 reporting sources 2 hours agoNew

Understand the event

AI overview

OpenAI shared findings from an assessment of its internal monitoring system by METR and Redwood Research, saying the findings helped strengthen safeguards as it expanded the system. OpenAI said it invited the organizations to conduct the evaluation over the summer and gave them deep access to its infrastructure for three weeks. METR said two employees from the organizations worked on premises to assess monitoring for misalignment in OpenAI’s internal agents, and that a brief write-up is now public.

Generated by AI from reports · Updated 1 hours ago

Latest development10/22 23:57 (UTC+8)
OpenAI shares embedded monitoring assessment by METR and Redwood ResearchOpenAI · X

Read next

More changes in today’s briefingMicrosoft Research and Collaborator Resolve the KLS Conjecture With Human-Directed AI

Report timeline

Follow the reports to explore different sides of the event.

10/22
  1. OpenAI · X
    OpenAI shares embedded monitoring assessment by METR and Redwood Research

    OpenAI said it invited METR and Redwood Research to evaluate its internal monitoring system this summer, providing deep infrastructure access over three weeks. OpenAI said the findings helped strengthen safeguards as it expanded the system. In the quoted post, METR said two employees from the organizations worked on premises to assess monitoring for misalignment in internal agents, and a brief write-up is now public. https://t.co/F9DSyjEOr1

Keep up with AI, one email a day

Model releases, price changes and compute developments, with original sources. Free, confirmed by email, and easy to unsubscribe. Read a past issue

Event heat trend

There is not enough continuous observation data to plot a trend yet.