Performance review season is one of the most time-consuming periods in any HR calendar. Writing balanced, specific, useful feedback for every employee — while managing everything else — is genuinely hard.
ChatGPT (and other AI tools like Claude and Gemini) can cut the time you spend on reviews significantly. Not by replacing your judgement, but by handling the structure and first draft so you can focus on making it accurate and fair.
Here's exactly how to use it.
## What AI can and can't do for performance reviews
AI is excellent at: structuring feedback, suggesting language for difficult observations, generating goal frameworks, and turning bullet points into polished paragraphs.
AI cannot: know what the employee actually did, replace your professional judgement, or write something genuinely personalised without your input.
The key is giving it enough context. The more you put in, the more useful the output.
## Step 1: Turn your notes into a structured draft
Most managers go into review season with rough notes — a few bullet points, some memory of what went well, a vague sense of what needs to improve.
Start there. Give ChatGPT your notes and ask it to structure them:
**Prompt:**
"I need to write a performance review for [name], our [job title]. Here are my notes: [paste your bullet points]. Turn these into a structured review with sections for: overall performance, key strengths (with specific examples), areas for development, and goals for the next review period. Tone should be direct, constructive, and encouraging. About 300 words."
You'll get a solid first draft in seconds. Edit it to make sure every specific detail is accurate — AI fills in gaps based on what sounds right, not what actually happened.
## Step 2: Write balanced feedback for difficult situations
The hardest reviews to write are for employees who are underperforming, or who have both strong and weak areas. It's easy to either soften the message too much or come across as overly critical.
**Prompt for an underperformer:**
"Write a performance review section for an employee who has not met their targets this quarter. The main issues are [list issues]. I want to be honest and direct without being harsh. The employee has shown [any positives]. Include a clear development plan with 2–3 specific actions and a timeline."
**Prompt for a mixed performer:**
"Write a balanced performance review for someone who excels at [strengths] but struggles with [weaknesses]. I want the review to acknowledge the genuine strengths, address the development areas honestly, and end on a constructive note. Avoid generic phrases like 'great team player'."
## Step 3: Set meaningful goals for the next period
Vague goals lead to vague outcomes. Use AI to build SMART goals that are actually tied to the employee's role.
**Prompt:**
"Create 3 SMART performance goals for a [job title] for the next 6 months. Current focus areas: [list 2–3 areas]. Each goal should include: what success looks like, how it will be measured, and a target date. Make them challenging but realistic."
## Step 4: Prepare for the review conversation
Writing the review is only half the job. The conversation is where it actually lands — or doesn't.
**Prompt:**
"Help me prepare for a performance review conversation with [name]. The review covers: [summary of key points]. I want to: open in a way that puts them at ease, deliver the development feedback clearly without it feeling like an attack, and close with clear next steps. Give me an outline of the conversation with suggested language for each stage."
## How much time does this actually save?
Most HR managers and people leaders report spending 45–90 minutes per review when writing from scratch. With AI handling the structure and first draft, that typically drops to 15–25 minutes — including editing and personalising the output.
Across 20 reviews, that's a full working day back.
## Which AI tool should you use?
All of the major tools work well for this: ChatGPT (OpenAI), Claude (Anthropic), Gemini (Google), and Microsoft Copilot. The free tiers are enough for review season.
The prompts in this article work across all four. The only difference is how you access them — the approach is the same.
---
Performance reviews are one of the highest-value places to use AI in HR. The time saving is immediate and the quality of output — when you give the AI enough context — is genuinely good.
If you want a complete set of performance review prompts (plus recruitment, onboarding, policy writing, and more), the AiVarise HR AI guides cover every core HR function with ready-to-use prompts for every scenario.