97 forecasters

Will leading AI labs have their models evaluated for dangerous behavior before 2026? (Google DeepMind)

20%chance

This question is closed for predictions.

Authors:

Opened:

Nov 18, 2025

Closed:

Nov 18, 2025

Scheduled resolution:

Jan 1, 2026

Spot Scoring Time:

Nov 18, 2025

Community Prediction Aggregation Method:Unweighted

Question Weight:69%

Include Bots:Yes