Identify every risk your AI product carries. Rate each one. Document how you fixed it. This protects you legally if something goes wrong.
| Risk Description | Likelihood | Impact | Risk Score | Mitigation Plan | Status |
|---|---|---|---|---|---|
| AI gives wrong medical advice to users | High | High | 9/10 | Added disclaimer on every AI response | Resolved |
| AI shows biased results for female job applicants | Medium | High | 7/10 | Tested model on 500 sample profiles for bias | In Progress |
| User data sent to OpenAI API without consent | Low | High | 6/10 | Added consent checkbox before AI features activate | Resolved |