c/ai-in-hr-people-analytics — AI in HR & People Analytics
-
Guardrails we added before letting AI draft review summaries
Managers were spending hours turning scattered feedback into a coherent narrative, so we let an assistant draft the first pass. It only became trustworthy once we added rules:
- It only summarises feedback that's already been given — never invents examples
- Every claim cites which peer or source said it
- The manager has to edit, not just approve, before it goes to the employee
-
Does anyone check AI-generated people-analytics outputs for bias before presenting them?
We're starting to use AI for workforce reporting and I keep wondering whether the outputs are quietly encoding bias from the underlying data. Is anyone actually auditing this, or is it still mostly trust-and-hope?