Home · Business · HR & People · Policies & handbooks

PARTLY

As of 13 August 2026, AI can only partly write a performance management policy.

This still needs a person who signs their name to it.

Can you do it?

5 minutesto a draft.

1 hourto something you’d act on.

Cost, all in£0

Skill neededchat-fluent

Who has to check ita professional

What the alternative costsThe supplied tool data does not give a price for an employment solicitor or HR consultant.

If this goes wrong: your managers apply an unclear or inconsistent process and the business faces a grievance, discrimination claim or dispute about dismissal.

What to actually do

  1. Hand it to a person

    The route this page recommends

    Someone with a licence or accountable authority has to sign this before it counts.

  2. Use a tool built for this

    Second choice

  3. Do it yourself

    The distant third

    A chat interface, chat-fluent skill, and roughly 1 hour until you can act on the result.

    How to actually do it

    1. Open your current staff handbook, employment contracts, disciplinary or capability procedures, sickness absence policy and equality or reasonable-adjustment guidance.
    2. Gather the business name, workforce details, manager roles, review cycle, existing appeal route and any differences between sites or employee groups.
    3. Paste the prompt and the gathered information into a chatbot, then paste the relevant policy text below it where the prompt asks for existing documents.
    4. Ask the chatbot to produce the draft, its missing-information list and a separate list of legal and operational points requiring review.
    5. Compare every proposed process step, timescale, entitlement and responsibility with your current contracts, handbook and the current employment guidance on GOV.UK.
    6. Send the draft and the flagged points to an employment solicitor or suitably qualified HR professional for review before adoption.
    7. Update the policy from the review, obtain approval from the person responsible for employment decisions, then publish the final version with an owner and review date.

    Prompt

    Draft a performance management policy for a UK business using the information below.

Business name: [BUSINESS NAME]
Business type and sector: [SECTOR]
Number of employees: [NUMBER]
Locations and working arrangements: [LOCATIONS AND REMOTE OR HYBRID DETAILS]
Employee groups covered: [EMPLOYEE GROUPS]
Who manages performance: [ROLES]
Review cycle: [REVIEW FREQUENCY]
Existing processes or documents: [DETAILS OR NONE]
Desired tone: plain, respectful and practical

The policy must cover purpose, scope, responsibilities, objectives, regular reviews, feedback, support and training, records, handling underperformance, reasonable adjustments, sickness absence where relevant, formal capability stages, employee representation, appeals, confidentiality and policy review. Distinguish informal support from formal capability action. Do not invent legal duties, deadlines, contractual terms or company procedures. Do not make decisions about a particular employee. Flag every point that needs checking against current UK employment law, our contracts or our existing handbook. Identify where an employment solicitor should review the draft. Use headings, short paragraphs and a table showing who does what at each stage. Begin with a short list of information still needed. State that the policy is a draft and not professional advice.

    Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.

What it gets wrong

  • AI cannot decide what performance standard is reasonable for each role or how much support a particular employee needs.
  • AI cannot know whether the draft conflicts with your contracts, established practice, collective arrangements or previous management decisions unless you provide and interpret that context.
  • AI cannot reliably resolve the overlap between performance, misconduct, sickness absence, disability and discrimination in a difficult case.
  • AI cannot take responsibility for a manager's decision to start formal action, dismiss an employee or respond to an appeal.
  • AI cannot replace a legal review of a policy that may later be relied on in an employment dispute.

What caps this at PARTLY: legal accountability, judgement under ambiguity and context depth.

How we scored this

Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.

AxisScore (0–2)
Output2
Inputs1
Verification1
Liability1
Effort delta2
Total7 / 10

The methodology and its thresholds are published in full.

FAQ

Can AI write a performance management policy?
Partly. AI can produce a useful first draft and organise the process, but you must supply your business rules and have the result checked for UK employment law, contractual conflicts and fair application.
Can I use an AI-generated performance policy in my UK business?
You can use it as a draft, not as an unreviewed final policy. The policy needs approval from a responsible person in your business, and a serious or unusual case should be checked by an employment solicitor.
What should a UK performance management policy include?
It should explain objectives, regular feedback, support, records, reasonable adjustments, informal improvement steps, formal capability stages, representation, appeals and confidentiality. It should also say who makes each decision and distinguish performance concerns from misconduct and absence matters.
Do I need a solicitor to write a performance management policy?
Not always, but legal review is sensible before you adopt a policy or use it in a serious case. This is not professional advice; an employment solicitor should review the policy where dismissal, disability, discrimination, contractual terms or an existing dispute may be involved.

Nearby answers

All Policies & handbooks rulings

Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.

The newsletter

AI news, new answers and product picks, straight to your inbox.