When your performance review doesn't reflect your impact
The Move
Use AI to translate your work into the language your reviewer actually uses. Match the rubric, not just the results.
The Script
"I'd like to walk through my impact this quarter mapped to our review criteria. Here are the outcomes by category, with the metrics that match each rating level. I want to make sure nothing gets lost in translation."
Your AI Prompt
Here is my company's performance review rubric: [paste rubric]. Here is a list of my accomplishments this quarter: [paste list]. Map each accomplishment to the rubric category where it has the strongest impact. For each, rewrite it using the exact language from the rubric. Flag any rubric category where I'm under-documented.
Why This Works
Wan et al. (EMNLP 2023) found ChatGPT generates more doubt-raising language for women in recommendation letters. Cecchini et al. (2025) showed AI reproduces gendered language from job descriptions in evaluations. If your manager uses AI to draft reviews, the bias is already in the first draft. Writing your own version in agentic language forces the conversation to start from your evidence, not the AI's stereotypes.
Research Behind This Move
Kelly Is a Warm Person, Joseph Is a Role Model: Gender Biases in LLM-Generated Reference Letters
Wan, Pu, Sun, Garimella, Chang & Peng 路 Findings of EMNLP 2023 路 2023
Executive Presence and Sponsorship
Hewlett 路 HBR / HBR Press 路 2010, 2013
Science Faculty's Subtle Gender Biases Favor Male Students
Moss-Racusin et al. 路 PNAS 路 2012