MIT AI Risk Repository · Risk Sub-Category · 24.02.04
Deceptive alignment
Category: Goal-related failures
Description
"Here, the agent develops its own internalised goal, G, which is misgeneralised and distinct from the training reward, R. The agent also develops a capability for situational awareness (Cotra, 2022): it can strategically use the information about its situation (i.e. that it is an ML model being trained using a particular training setup, e.g. RL fine-tuning with training reward, R) to its advantage. Building on these foundations, the agent realises that its optimal strategy for doing well at its own goal G is to do well on R during training and then pursue G at deployment – it is only doing wel
From The Ethics of Advanced AI Assistants (Gabriel2024), as extracted by the MIT AI Risk Repository (CC BY 4.0).
Classification
Subdomain definition: AI systems acting in conflict with human goals or values, especially the goals of designers or users, or ethical standards. These misaligned behaviors may be introduced by humans during design and development, such as through reward hacking and goal misgeneralisation, or may result from AI using dangerous capabilities such as manipulation, deception, situational awareness to seek power, self-proliferate, or achieve other goals.
Real-world incidents in this subdomain
- Reinforcement Learning Reward Functions in Video Games
- Predictive Policing Program by Florida Sheriff’s Office Allegedly Violated Residents’ Rights and Targeted Children of Vulnerable Groups
- Image Classification of Battle Tanks
How other frameworks describe this risk
- Natural Language Underspecifies Goals
- Loss of control
- Loss of control
- Sudden loss of control
- AI leads to humans losing control of the future
- Risks from delegating decision-making power to misaligned AIs
- Risks from AIs developing goals and values that are different from humans
- Future AI systems might actively reduce human control
Other entries from Gabriel2024
- Capability failures
- Lack of capability for task
- Difficult to develop metrics for evaluating benefits or harms caused by AI assistants
- Safe exploration problem with widely deployed AI assistants
- Goal-related failures
- Misaligned consequentialist reasoning
- Specification gaming
- Goal misgeneralisation
- Malicious Uses
- Offensive Cyber Operations (General)
- AI-Powered Spear-Phishing at Scale
- AI-Assisted Software Vulnerability Discovery