The AI Harm & Stakeholder Map
Template
A working canvas for identifying who could be harmed by an AI system, what type of harm is realistic, and how severe it could be.
Stakeholders and affected people
- List everyone who interacts with or is affected by the system, not just direct users.
- Identify anyone in a group that could be disproportionately affected.
Type and severity of harm
- Classify the harm as individual, group, organisational or societal.
- Rate probability and severity for each realistic harm scenario.
- Flag any harm severe enough that it should block launch on its own.
Mitigation hierarchy
- For each harm, decide: avoid, reduce, transfer/share, accept or monitor.
- Record who owns the decision and when it will be reviewed again.
Start a conversation
Have a product or AI decision to make?
Useful first calls usually start with one unclear decision, a deadline and a team that needs a practical next move.
Tell us about it