EveryAILaw.comIncident Reporting
Requirement to report AI-related incidents, safety concerns, or failures to regulatory authorities within specified timeframes.
What Counts
- Mandatory incident reporting to regulators
- Whistleblower protection for AI safety concerns
- Defined reporting timelines (e.g., 72 hours)
- Serious incident notification requirements
- Post-incident analysis documentation
What Does Not Count
- Voluntary incident sharing
- Internal-only incident tracking without regulatory reporting
- General product liability reporting without AI-specific provisions
Related Terms
- Safety reporting
- Breach notification
- Whistleblower protection
- Post-market surveillance
Implementing Regulations
| Regulation | Jurisdiction | Status | Provisions |
|---|---|---|---|
| California SB 53 (Frontier AI Transparency Act) | California | enforcing | 1 |
| Interim Measures for Generative AI Services | China | enforcing | 1 |
| Digital Operational Resilience Act (DORA) | European Union | enforcing | 1 |
| New York RAISE Act | New York | enacted | 1 |
Standards & Frameworks
Voluntary standards that address this obligation. Implementing these can help demonstrate compliance with the binding regulations above.
| Standard | Type | Status | Provisions |
|---|---|---|---|
| Hiroshima AI Process – Principles & Code of Conduct | framework | voluntary | 1 |