Infrastructure & Risk
AI Safety Incidents
A public-source incident dashboard framework for AI harms, near-harms, risk categories, verification status, and source reports.
How to use this dashboard
A public-source incident dashboard framework for AI harms, near-harms, risk categories, verification status, and source reports.
Use this dashboard to review AI incident records, harm categories, sectors, countries, and verification status before citing an incident.
AI Safety Incidents
5 records| Data source | AI Incident Database snapshot feed | Multiple AI systems | Cross-sector | Harm / near-harm reports | Global | Source-defined / manual review | Use snapshot rows before publication |
| Data source | MIT AI Risk Repository incident tracker | Multiple AI systems | Cross-sector | Risk taxonomy classifications | Global | Taxonomy-based | Classification reference source |
| News signal | AI incident news watch query | Models / apps / automated systems | News-monitored sectors | Reported allegations, failures, harms, or investigations | From article metadata when available | Manual review needed | News signal, not confirmed incident |
| Risk source | AI risk taxonomy reference | General AI systems | Governance / risk management | Risk categories and causal pathways | Global | Not an incident feed | Use to classify incident rows |
| Manual intake | Editor-reviewed incident candidate | To be filled from source report | To be classified | To be classified | To be verified | Pending review | Manual queue |