Work · what HR is reading
AI in performance reviews
30-second gist~30s read
AI summarisers are creeping into performance review prep. Your manager opens an HR tool, it generates a year-in-review of your work — the meetings you attended, the things you said, the projects you owned. Some of this is helpful. Some of it surfaces patterns you wouldn't have noticed. All of it depends on what data the AI was given.
Worth understanding what's in the picture before performance season catches you by surprise.
If you want more
What gets pulled in
- Meeting transcripts — anything captured by Teams, Zoom, or Google Meet recording.
- Calendar events — frequency, duration, attendee patterns.
- Project tracker entries — Jira, Asana, Trello completion data.
- Sometimes Slack/Teams messages — depends on the company's policy and contract.
- Sometimes code commits — for engineering roles.
Notably not usually included: your private email, files in personal drives, conversations outside the work tools.
What you can ask
At most companies you can ask HR what data sources the AI used and what it scored you on. Some HR tools now generate a candidate-facing summary on request. If your review surfaces a pattern you don't recognise, that's a fair thing to push back on — "can I see the underlying examples the AI flagged?".
Treat the AI summary as a draft, not a verdict. The same applies to your manager.