THE TAKEAWAY
Start with a decision and its evidence needs, not with the number of reports available.
My assessment: an AI summary earns its place only if it helps a reviewer ask a better question and trace the answer back to the evidence.
- Where AI could help
- Explore AI-assisted report triage, theme discovery and comparisons across periods, with sensitive data protected and source passages retained.
- What remains uncertain
- A model can mistake reporting changes for a risk trend. We have not tested a model on a representative occurrence-report dataset.
- The human decision
- A qualified reviewer validates coding, exposure and alternative explanations before changing a risk assessment.
- Evidence & confidence
- ICAO supports the safety-intelligence discipline; the suggested AI methods and safeguards are our interpretation, not ICAO-prescribed tools.
AI-generated interpretation · Prepared 16 September 2026 · Professionally reviewed and approved 16 September 2026. Sources are listed below; no live operational data are connected.
A more useful question
ICAO’s Safety Intelligence Manual, Doc 10159, addresses the collection, processing, analysis and use of safety data and information. It complements Doc 9859 and supports Annex 19 Amendment 2. Source ↗
Our editorial position is simple: intelligence should change the quality of a decision. More charts are not necessarily more understanding. Ask what the decision-maker knows now that they did not know before, and what uncertainty remains.
A hypothetical runway example
Suppose an airline identifies more unstable approaches at one destination. Before calling it a worsening safety trend, investigate exposure, fleet mix, weather, reporting coverage and any changed event definition. A rising count can represent changed activity or better detection as well as increased risk.
The aerodrome and operator can then compare relevant observations using agreed definitions. Possible questions include whether events cluster around particular conditions, whether a control is working as intended, and which additional evidence would distinguish competing explanations. This example is a discussion aid, not a finding about a real airport.
Make the reasoning inspectable
A useful briefing separates observation, interpretation and proposed action. Show the period covered, the population observed, important exclusions and the owner of the next decision. Keep a record of why an intervention was chosen.
When a new data source changes the picture, revisit the conclusion visibly. An analytical process that can explain its own correction is more credible than one that presents every output as certain.
A QUESTION FOR YOUR NEXT SAFETY DISCUSSION
Can your latest safety dashboard identify a decision that changed because of its evidence?
Follow the evidence
Sources checked for this edition on 16 September 2026. Uncited examples and recommendations are IASMS editorial analysis, not official requirements. AI-produced editorial analysis reviewed by an experienced aviation safety professional for aviation-safety context, evidence boundaries, accountability and publication suitability. This is not regulatory, legal or independent specialist certification. How we work →