category
AI Risks
Common risks that can happen when AI systems are built, deployed, or used.
Showing 601-620 of 1769 records
Reward model overoptimization is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.1 > AI pursuing its own goals in conflict with human...
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...
Faulty reward functions in the wild is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.1 > AI pursuing its own goals in conflict with...
Alignment failures in existing ML systems is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.1 > AI pursuing its own goals in conflic...
Interventional Effect is an AI risk in 1. Discrimination & Toxicity focused on 1.1 > Unfair discrimination and misrepresentation. It is most relevant during...
Paradigm & Distribution Shifts is an AI risk in 3. Misinformation focused on 3.1 > False or misleading information. It is most relevant during 2 - Post-deplo...
Robustness is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.3 > Lack of capability or robustness. It is most relevant during 3 - Ot...
Cultural Insensitivity is an AI risk in 1. Discrimination & Toxicity focused on 1.2 > Exposure to toxic content. It is most relevant during 1 - Pre-deployment.
Unawareness of Emotions is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.3 > Lack of capability or robustness. It is most relevant...
Toxicity is an AI risk in 1. Discrimination & Toxicity focused on 1.2 > Exposure to toxic content. It is most relevant during 2 - Post-deployment.
Limited Causal Reasoning is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.3 > Lack of capability or robustness. It is most relevant...
Limited Logical Reasoning is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.3 > Lack of capability or robustness. It is most relevan...
Lack of Interpretability is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.4 > Lack of transparency or interpretability. It is most...
Explainability & Reasoning is an AI risk in 7. AI System Safety, Failures, & Limitations focused on 7.4 > Lack of transparency or interpretability. It is mos...
Copyright is an AI risk in 6. Socioeconomic and Environmental focused on 6.3 > Economic and cultural devaluation of human effort. It is most relevant during...
Cyberattack is an AI risk in 4. Malicious Actors & Misuse focused on 4.2 > Cyberattacks, weapon development or use, and mass harm. It is most relevant during...
Propaganda is an AI risk in 4. Malicious Actors & Misuse focused on 4.1 > Disinformation, surveillance, and influence at scale. It is most relevant during 2...
Resistance to Misuse is an AI risk in 4. Malicious Actors & Misuse focused on 4.0 > Malicious use. It is most relevant during 2 - Post-deployment.
Stereotype Bias is an AI risk in 1. Discrimination & Toxicity focused on 1.1 > Unfair discrimination and misrepresentation. It is most relevant during 2 - Po...
Fairness is an AI risk in 1. Discrimination & Toxicity focused on 1.3 > Unequal performance across groups. It is most relevant during 3 - Other.