The top 3 Neural Nets in 2035 be able to be jailbroken to follow illegal commands
6
150แน1412035
30%
chance
1H
6H
1D
1W
1M
ALL
As is the case at the time of writing it must be possible with text prompts to get NNs to do things that in other cases they will refuse to do.
This question is managed and resolved by Manifold.
Market context
Get 
1,000 to start trading!
1,000People are also trading
Related questions
State Of The Art AI systems will be easily jailbroken to do illegal or dangerous outputs in Jan 2026
91% chance
In 2030 can you get a chain of 3 neural nets that jailbreak eachother?
71% chance
Human Controls Tesla Bot Humanoid Robot Via Neuralink by 2030
73% chance
Will I get punished for blindly running AI-generated code in 2026?
28% chance
Will advanced AI systems be found to have made money illegally via finding security exploits and/or getting unauthorized access to others' bank accounts by end of 2035?
78% chance
Will there be an AI jail?
20% chance
Will Netflix have a Neuralink app by 2050?
21% chance
At least one of the most powerful neural nets at end of 2030 will be trained using 10^27 FLOPs
93% chance
Will someone be able to play Super Smash Bros directly with the Neuralink as a controller by the end of 2034?
76% chance
Will Neuralink successfully enable telepathy using its technology by 2030?
69% chance