By what year will any AI lab release a model under ASL-4 measures?
3
625แน752030
42%
2026
50%
2027
50%
2028
54%
2029
58%
2030
I'll be using AI safety levels as defined by Anthropic. Anthropic had released all models at ASL-2 until Opus 4, which is ASL-3. By what year will we see a lab using ASL-4 measures?
Note that ASL-4 isn't even defined yet - I will resolve based on whatever standard Anthropic ends up publishing. If Anthropic goes bankrupt / doesn't publish for a long time and someone else releases their own ASL-4 standard then I'll discuss with the market whether to use that standard, but ultimately it will come down to my inscrutable whims.
Time ranges are upper bounds - for 2026 to resolve YES it must happen before 01/01/2026.
Answers resolve inclusively - if 2029 resolves YES then so does 2030.
This question is managed and resolved by Manifold.
Get 
1,000 to start trading!
1,000People are also trading
Related questions
Will a new lab create a top-performing AI frontier model before 2028?
93% chance
Will Anthropic announce one of their AI systems is ASL-4 or higher before the end of 2025?
7% chance
When will there first be an AI system that qualifies as ASL-3?
When will there first be a credible report that an AI system qualifies as ASL-3?
Will Anthropic announce one of their AI systems is ASL-3 before the end of 2025?
89% chance
When will Anthropic first train an AI system that they claim qualifies as ASL-3?
Limits on AI model size by 2026?
15% chance
Will it be public knowledge by EOY 2025 that a major AI lab believed to have created AGI internally before October 2023?
7% chance
In what year will an AI Lab in China build Artificial General Intelligence?
2031
Will models be able to do the work of an AI researcher/engineer before 2027?
26% chance