52 forecasters

Will the listed AI companies/labs pause any of their models before January 1st, 2025 because they detect dangerous capabilities?

Anthropicresult: No

OpenAIresult: No

Google DeepMindresult: No

Anthropic

OpenAI

Google DeepMind

Authors:

jleibowich will_aldred

Opened:

Sep 30, 2023

Closes:

Dec 31, 2024

Resolved:

Jan 13, 2025

Artificial Intelligence

Computing and Math

🏆 2021-2025 Leaderboard

Newest generation of GPT models show major drop in political bias, OpenAI says

New York Post•Oct 10, 2025

ChatGPT safety systems can be bypassed to get weapons instructions

NBC•Oct 10, 2025

Anthropic's latest AI model can tell when it's being evaluated: 'I think you're testing me'

Business Insider•Oct 7, 2025

Learn more about Metaculus NewsMatch

Will OpenAI, Google DeepMind, or Anthropic announce that they are pausing all training runs above a certain size for safety reasons, before 2026?

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

Google DeepMind

Will the CEO of OpenAI, Meta, or Alphabet (Google) publicly commit to specific limitations on their company’s AI system autonomy before January 1, 2027?