COMPEL Glossary / risk-appetite
Risk Appetite
Risk appetite is the overall level and types of risk that an organization is willing to accept in pursuit of its strategic objectives, set by the board of directors or equivalent governing body.
What this means in practice
For AI, risk appetite statements define boundaries such as acceptable model accuracy thresholds, tolerable bias levels, approved use case categories, and permitted autonomy levels for AI systems. Risk appetite is broader than risk tolerance (which applies to specific risks) and serves as the foundational governance parameter that shapes all downstream risk management decisions. For organizations, a clearly articulated AI risk appetite prevents both excessive caution (rejecting all AI projects due to undefined risk concerns) and reckless deployment (launching AI without adequate safeguards). In COMPEL, risk appetite setting is part of the governance architecture designed in Module 3.4, Article 5.
Why it matters
A clearly articulated AI risk appetite prevents both excessive caution that rejects all AI projects and reckless deployment without adequate safeguards. Risk appetite set at the board level defines boundaries for acceptable model accuracy, bias levels, use case categories, and AI autonomy levels. Without it, every AI decision becomes a new debate about acceptable risk, slowing transformation and creating inconsistent governance across the organization.
How COMPEL uses it
Risk appetite setting is part of the governance architecture designed in Module 3.4, Article 5 during the Model stage. During Calibrate, the current risk posture is assessed against the organization's stated appetite. The Governance pillar translates risk appetite into operational risk management criteria. The Evaluate stage verifies that deployed AI systems operate within defined risk boundaries, and the Learn stage refines appetite based on observed outcomes.
Related Terms
Other glossary terms mentioned in this entry's definition and context.