category
AI Risks
Common risks that can happen when AI systems are built, deployed, or used.
Showing 1001-1020 of 1769 records
Violated expectations is an AI risk in 5. Human-Computer Interaction focused on 5.1 > Overreliance and unsafe use. It is most relevant during 2 - Post-deploy...
Self-Actualisation Harms is an AI risk in 5. Human-Computer Interaction focused on 5.2 > Loss of human agency and autonomy. It is most relevant during 2 - Po...
Sociocultural and Political Harms is an AI risk in 5. Human-Computer Interaction focused on 5.2 > Loss of human agency and autonomy. It is most relevant duri...
Economic Harms is an AI risk in 6. Socioeconomic and Environmental focused on 6.2 > Increased inequality and decline in employment quality. It is most releva...
Authoritarian Surveillance, Censorship, and Use (General) is an AI risk in 4. Malicious Actors & Misuse focused on 4.1 > Disinformation, surveillance, and in...
Harmful Content Generation at Scale (General) is an AI risk in 4. Malicious Actors & Misuse focused on 4.1 > Disinformation, surveillance, and influence at s...
Malicious Code Generation is an AI risk in 4. Malicious Actors & Misuse focused on 4.2 > Cyberattacks, weapon development or use, and mass harm. It is most r...
AI-Assisted Software Vulnerability Discovery is an AI risk in 4. Malicious Actors & Misuse focused on 4.2 > Cyberattacks, weapon development or use, and mass...
Offensive Cyber Operations (General) is an AI risk in 4. Malicious Actors & Misuse focused on 4.2 > Cyberattacks, weapon development or use, and mass harm. I...
Deceptive alignment is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.1 > AI pursuing its own goals in conflict with human goals or...
Goal misgeneralisation is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.1 > AI pursuing its own goals in conflict with human goals...
Specification gaming is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.1 > AI pursuing its own goals in conflict with human goals or...
Misaligned consequentialist reasoning is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.3 > Lack of capability or robustness. It is...
Goal-related failures is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.1 > AI pursuing its own goals in conflict with human goals o...
Safe exploration problem with widely deployed AI assistants is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.3 > Lack of capability...
Difficult to develop metrics for evaluating benefits or harms caused by AI assistants is an AI risk in 6. Socioeconomic and Environmental focused on 6.5 > Go...
Capability failures is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.3 > Lack of capability or robustness. It is most relevant duri...
Self-improvement is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.2 > AI possessing dangerous capabilities. It is most relevant dur...
Situational awareness, for instance if this causes a model to act differently in training compared to deployment, meaning harmful characteristics are missed...
The ability to cooperate with other highly capable AI systems is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.2 > AI possessing da...