The likelihood that a harm will occur and the severity: What two dimensions does this matrix combine to help

A probability/severity harms matrix ranks risks by combining how likely a harm is with how severe its impact would be.

The question

Preparing for AIGP? Take the free 5-min readiness quiz →

1. The cost of a control and the time required to implement it, combined to schedule mitigation work efficiently.

   Plausible but wrong: cost and schedule inform project planning, not the likelihood-and-severity axes of a harms matrix.
2. The likelihood that a harm will occur and the severity of its impact if it does, combined to rank which risks to address first. ✓

   Correct: a probability/severity matrix ranks risks by combining how likely a harm is with how severe its impact would be.
3. The accuracy of the model and the size of its training dataset, combined to estimate how reliable its predictions will be in use.

   Plausible but wrong: accuracy and data size are performance measures, not the two risk dimensions the matrix combines.
4. The number of affected users and the geographic regions served, combined to size the overall market reach of the deployed system.

   Plausible but wrong: reach and geography describe scale of use, not the probability and severity that the harms matrix plots.

How to remember it

A probability/severity harms matrix ranks risks by combining how likely a harm is with how severe its impact would be.

More Understanding How to Govern AI Development questions

• Stakeholder engagement and feedback: Which best practice does this activity represent? →
• It establishes compliance and supports risk management: Which statement best captures the primary governance →
• Whether the data is fit-for-purpose and representative: Under data governance requirements, which concern →
• All 426 Understanding How to Govern AI Development questions →