By what year will any AI lab release a model under ASL-4 measures?
6
Ṁ3052030
1D
1W
1M
ALL
51%
2026
57%
2027
61%
2028
65%
2029
68%
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.
Get Ṁ1,000 play money
Related questions
Related questions
Will Anthropic announce one of their AI systems is ASL-4 or higher before the end of 2025?
5% chance
Will Anthropic announce one of their AI systems is ASL-3 before the end of 2025?
93% chance
Limits on AI model size by 2026?
5% chance
When will Anthropic reach or surpass ASL-4 [metaculus]
Will AI (large language models) collapse by may 2026?
11% chance
OpenAI's next "GPT-OSS" release date
Will a new lab create a top-performing AI frontier model before 2028?
87% chance
By what date will all state-of-the-art general-purpose AI systems not be reasoning models?
By what date will at least one state-of-the-art general-purpose AI system not be a reasoning model?
In what year will an AI Lab in China build Artificial General Intelligence?
-