AI Risk Domains
Explore categorized AI risks derived from the AI Risk Repository.
Discrimination & Toxicity
Risk coverage: 14%
- Unfair discrimination and misrepresentation
- Exposure to toxic content
- Unequal performance across groups
Malicious Actors & Misuse
Risk coverage: 16%
- Disinformation and influence at scale
- Cyberattacks and mass harm
- Fraud and scams
Socioeconomic & Environmental
Risk coverage: 19%
- Power centralization
- Inequality & employment decline
- Environmental harm
AI System Safety, Failures, & Limitations
Risk coverage: 25%
- Misalignment with human values
- Dangerous capabilities
- Lack of robustness