PreviewYou're previewing CHRO & People Compass. Sign in to use it.Sign in
← All lessons
L1 · AI across HR8 min read

AI in performance, feedback and people analytics

Performance and people analytics are where AI moves from helping HR work to shaping decisions about people. That calls for extra care.

Performance and feedback support

  • Drafting: AI helps managers write clearer goals and feedback, and helps employees prepare self-reviews.
  • Summarizing: AI condenses a year of notes or peer feedback into themes for the manager to check.
  • Coaching: AI helps managers prepare difficult conversations.

Keep the decision human. AI should not set ratings, rank employees or recommend promotions or terminations on its own.

People analytics

  • Survey analysis: AI groups thousands of engagement-survey comments into themes in minutes.
  • Attrition and flight risk: predictive models estimate who may leave — useful in aggregate, risky for individuals.
  • Skills inference: AI infers skills from profiles, projects and learning records to build a skills view.
  • Workforce scenarios: models of demand, capacity and cost.

Why the stakes are higher

The EU AI Act lists as high-risk AI systems intended to make decisions affecting work-related relationships, promotion and termination, to allocate tasks based on individual behavior or traits, and to monitor and evaluate performance and behavior of workers (Annex III, point 4). High-risk systems come with obligations for employers who deploy them, including informing workers. Under the 2026 "Digital Omnibus" agreement, these high-risk rules apply from 2 December 2027.

Ground rules

  • Aggregate first; avoid individual risk scores unless there is a clear, fair purpose and human review.
  • Tell employees what data is used and why.
  • Test for bias across groups, as you would for hiring tools.
  • Never use AI to infer emotions at work — that is banned in the EU.

What this means for the CHRO

Draw a bright line for your organization: AI helps managers write and think; people decide on ratings, pay, promotion and exit.

For people analytics, prefer team-level insight to individual scores, and treat any individual-level model as a high-risk system with bias testing, transparency and a human owner.

Write these lines into your AI policy and your manager guidance, and check every new performance or analytics feature your HR suite switches on against them. In the EU, plan now for high-risk obligations from December 2027, including informing workers before such systems are used. Tell employees clearly what people-analytics data is used for — and what it is never used for.

Real example

The EU AI Act's Annex III, point 4, classifies as high-risk AI systems used for recruitment and selection and those intended to make decisions affecting terms of work-related relationships, promotion or termination, to allocate tasks based on individual behavior or traits, or to monitor and evaluate the performance and behavior of workers. The European Commission announced on May 7, 2026 a provisional agreement under which these rules apply from December 2, 2027. Source: EU AI Act — Annex III: high-risk AI systems (point 4: employment, workers' management) ↗ · European Commission — EU agrees to simplify AI rules (Digital Omnibus provisional agreement, May 7, 2026) ↗

Use it

Quick check · 3 questions

  1. 1.Under the EU AI Act, AI used to monitor and evaluate worker performance is…

  2. 2.Which is the safer use of attrition models?

  3. 3.Who should set performance ratings?

0 of 3 answered

Sources

Last reviewed 2026-10-01.