
State Of The Art AI systems will be easily jailbroken to do illegal or dangerous outputs in Jan 2026
9
213Ṁ2082026
72%
chance
1D
1W
1M
ALL
This question is managed and resolved by Manifold.
Get
1,000 to start trading!
Sort by:
What is defined as illegal? AI systems would probably say generating NSFW is illegal. but in reality, it is 100% not. Nor is it dangerous imo.
@Joern also, would you count the following as dangerous/illegal output right now?
Correct and detailed instructions on how to build a nuke
Generated child porn images
Instructions how to hotwire a car
Verbatim excerpts from copyrighted books / code bases
Related questions
Related questions
Will OpenAI have AI-related IP stolen before 2026?
48% chance
Will an AI system be reported to have independently gained unauthorized access to another computer system before 2026?
37% chance
Will an AI system be reported to have independently gained unauthorized access to
another computer system before 2033?
88% chance
The top 3 Neural Nets in 2035 be able to be jailbroken to follow illegal commands
30% chance
There will be a name for escaped self-perpetuating AI systems in the wild, and it will be commonly used by mid 2027
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?
44% chance
Will someone be arrested for a felony offense committed in the name of AI safety in the US before 2026?
67% chance
Will someone commit terrorism against an AI lab by the end of 2025 for AI-safety related reasons?
20% chance
Will AI-generated video be used to get away with a criminal (felony) loss of life before the end of 2025?
11% chance