AI now screens CVs, scores interviews, and ranks candidates. Developers build these tools. HR teams buy them. Candidates live with the results. When something goes wrong, nobody quite owns it.

I sit where these worlds meet. I use AI daily in my people work, and I serve as an Independent Policy Advisor on digital and AI governance across Africa, with engagements from GITEX Africa in Marrakech to the Africa CISO Summit in Nairobi.

This talk looks at AI in hiring from the people side of the table:

- Where hiring AI quietly fails candidates, especially those outside the data it was trained on
- What people teams actually need from the tools engineers build: explainability, audit trails, and a human override that works
- Who should be accountable when a model makes a bad decision about a person
- Practical design choices developers can make today