Question

Will any frontier model be released in 2029 without a third-party evaluation of dangerous capabilities?

2 comments2

Total Forecasters27

Community Prediction40% (37% - 65%)

Make a Prediction

50%

community: 40%

Authors:

sandman

Opened:

Jan 23, 2025

Closes:

Jan 1, 2030

Scheduled resolution:

Jan 15, 2030

AGI Readiness Index

Artificial Intelligence

Radical Data Science

•Feb 25, 2025

Anthropic launches a new AI model that ‘thinks’ as long as you want

TechCrunch•Feb 25, 2025

Claude 3.7 Sonnet: Anthropic unveils advanced AI hybrid reasoning model

CNN•Feb 25, 2025

Learn more about Metaculus NewsMatch

Will leading AI labs have their models evaluated for dangerous behavior before 2026?

Before 2028, will powerful open-source AI be regulated more tightly than closed-source AI, through newly-enacted US law?

70%

Will every frontier AI lab collaborate with at least one other frontier lab on AI alignment or safety research, on January 1, 2030?

18%

Comments

? comments