As of 13 August 2026, AI can only partly document an employee performance concern.

What to actually do

1. Hand it to a person

   The route this page recommends

   A person who owns the outcome does this end to end, worth it when the failure is dear.

2. Use a tool built for this

   Second choice

3. Do it yourself

   The distant third

   A chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.

   How to actually do it

   1. Open the relevant job description, objectives, performance policy and any reasonable-adjustment guidance, then gather the exact wording that applies.
   2. Collect dated notes, work examples, targets, emails and meeting records, separating what you personally observed from what someone else reported.
   3. Remove unnecessary personal data and replace the employee's name with initials before pasting material into the chatbot.
   4. Paste the supplied prompt and replace each bracketed slot with the gathered facts, leaving a slot as [MISSING] when you do not have reliable information.
   5. Ask the chatbot to produce the document and its verification checklist, then compare every factual statement, date and quotation against the original records.
   6. Ask your HR colleague to check whether the wording is fair, consistent with policy, sensitive to disability or other protected characteristics, and suitable for the intended process.
   7. Amend the document from those checks, give the employee a proper opportunity to respond, and store and share the final version through your approved confidential HR system.

   Prompt

   Draft a factual employee performance concern document for a UK workplace using only the information below. Do not invent facts, motives, quotations, dates, policy requirements or legal conclusions. Separate directly observed facts from reported information and unverified claims. Use neutral, specific language and avoid labels such as lazy, difficult or unsuitable. Include: employee role, relevant expectation or agreed objective, each concern with its date and evidence, the impact on work, the employee's explanation if known, support or reasonable adjustments already offered, the improvement required, a practical review date or period, and the employee's right to respond. If information is missing, mark it [MISSING] rather than guessing. Flag any wording that could be discriminatory, retaliatory, disproportionate or inconsistent with the supplied policy. Do not recommend dismissal or disciplinary action. End with a checklist of facts and sources I must verify before using the document. Treat all information as confidential and use initials or a fictional name in the draft where possible.
   
   Employee role: [ROLE]
   Relevant objectives, job description or policy: [PASTE]
   Observed events and dates: [PASTE]
   Evidence or records: [PASTE]
   Employee explanation: [PASTE]
   Support, training or adjustments offered: [PASTE]
   Known context that may affect the concern: [PASTE]
   Intended audience and document purpose: [PASTE]

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