Swiss cheese model
Reason’s model of accidents passing through weaknesses in multiple layers of defence.
ConceptSources & confidence: Qualified confidence
Checked 2 October 2026 · AI-assisted review
Checked the definition, attribution and stated distinctions against the listed source passages.
- James Reason, Human error: models and management (2000)
Primary article’s indexed sections on layered defences, active failures and latent conditions.
What this check doesn’t establish
- Primary article sections were available through indexed text; direct full-page access was blocked.
- No claim is made that the graphical model first appeared in a particular year before this 2000 account.
Confidence applies to the scope above, not every claim made by the underlying source. How confidence works
Reason’s model represents defences as layers with weaknesses. An accident becomes possible when a trajectory passes through the available gaps, including those created by both active errors and latent organisational conditions.
It is a model for thinking about interacting safeguards, not proof that every accident requires every conceivable safeguard to fail. The essay applies the analogy to AI deployment; that application is not itself a validated safety case.
Record history
Catalogue source: knowledge/concepts/swiss-cheese-model.md
Generated by claude-code/claude-opus-5 on 2026-07-26