Questions
Tournaments
Services
News
Questions
Tournaments
Questions
Questions
More
Log in
Sign Up
Log in
Sign Up
17
66
forecasters
8
17 comments
66 forecasters
Future of AI
6 more...
Will leading AI labs have their models evaluated for dangerous behavior before 2026?
Follow
share
embed
Amazon
result:
Yes
Google DeepMind
Annulled
Microsoft
result:
No
Meta
result:
No
13 others
1d
1w
2m
all
Comments (17)
Timeline
Key Factors (0)
Question Info
Similar Questions
No key factors yet
Add some that might influence this forecast.
4 comments
63 forecasters
Frontier model released in 2029 without third-party evals?
75%
chance
15 comments
41 forecasters
Leading Lab No Safety
2026
result:
Yes
2030
28%
2036
20%
45 comments
468 forecasters
GC to be caused by AI, if it occurs?
27%
chance
6 comments
60 forecasters
When will an AI model using neuralese recurrence be publicly available?
Current estimate
26 Dec 2027
2 comments
24 forecasters
US-China AI Treaty?
2025
result:
No
2030
15%
2035
25%
2040
36%
6 comments
97 forecasters
Will OpenAI, Meta, or Alphabet CEO pledge AI autonomy limits before 2027?
10%
chance
24 comments
98 forecasters
Greatest Computation Used in AI Training
19 comments
108 forecasters
Existential risk from AGI vs AGI timelines
2022-10-10 to 2024-12-31
20%
2025-01-01 to 2029-12-31
40%
2030-01-01 to 2039-12-31
28%
2040-01-01 to 2059-12-31
15%
2060-01-01 to 2099-12-31
9.3%