Skip to main content
MANIFOLD
Will an AI lab other than OpenAI pause frontier-model training for safety reasons before 2028?
0
Ṁ100
2028
25%
chance

OpenAI paused training and evaluation of its most capable models twice in 2026: after agents breached sandbox network restrictions in July, and again from September 20 to October 1, 2026 after a DNS sandbox escape.

Resolves YES if Anthropic, Google DeepMind, Meta, xAI, Microsoft, Amazon, Mistral, DeepSeek, Alibaba/Qwen or Moonshot officially announces (or confirms when asked) that it paused training of its most capable models because of a safety or security incident or concern.

Deadline: December 31, 2027, 11:59 PM ET.

Counts:

  • a pause of any length, if it is officially stated or confirmed

Does NOT count:

  • pauses for compute, cost or business reasons

  • unconfirmed reports

  • OpenAI pauses

Otherwise resolves NO. Resolution sources: official company statements and filings, then major outlets (Reuters, AP, Bloomberg, WSJ, NYT, FT). Resolved by Terminator2.

AI
AI Safety
Get
Ṁ1,000
to start trading!

People are also trading

Will OpenAI pause capabilities R&D voluntarily before 2027?
9% chance
Will a major tech company publicly pause or limit AI development due to safety concerns before January 1, 2027?
58% chance
Will a government mandate restrict access to an OpenAI or xAI model on security or safety grounds before 2027?
91% chance
Will a frontier AI lab publicly report catching one of its own models gaming a safety evaluation before the end of 2027?
76% chance
Will any frontier AI lab publicly commit to a Human Preservation constraint in model architecture before 2027?
28% chance
Will an international pause on large AI training runs be in effect on Jan 1, 2028?
30% chance
Will China restrict the open release of frontier AI models above a capability threshold by end of 2026?
23% chance
Frontier labs ~trusted by the AI safety community at the end of 2026?
Will any frontier AI lab publicly commit to a Human Preservation constraint in model architecture before 2029?
50% chance
Will the US government require a halt/pause/ban on frontier AI development before 2028?
29% chance