As of 13 August 2026, AI can only partly create a risk assessment template.
This still needs a person who signs their name to it.
Can you do it?
5 minutesto a draft.
30 minutesto something you’d act on.
Cost, all in£0
Skill neededchat-fluent
Who has to check ita colleague
What the alternative costsNo comparable UK alternative price is provided in the supplied tool data.
If this goes wrong: the template omits an important hazard or records an unsuitable control, leaving the responsible employer to deal with possible injury, enforcement or disruption.
What to actually do
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.
Use a tool built for this
Second choiceDo it yourself
The distant thirdA chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.
How to actually do it
- Open a blank document or spreadsheet and note the workplace, activity, location, people exposed and the person responsible for the assessment.
- Gather existing procedures, incident records, equipment instructions, substances information and details of current controls without including unnecessary personal data.
- Paste the prompt into ChatGPT, Claude or Gemini and replace the bracketed slots with the workplace or activity details.
- Copy the generated table into the document or spreadsheet, then add the actual hazards, possible harms, existing controls, responsible people and review arrangements.
- Walk through the activity with the workers who do it and compare each row with what happens in practice, adding hazards or controls the template missed.
- Check the likelihood, severity and risk ratings against the stated scoring method, and ask a competent health and safety colleague to challenge the controls and residual risks.
- Send the completed assessment to the responsible manager for approval, assign the recorded actions and set the review date or trigger.
Prompt
Create a UK workplace risk assessment template for [workplace or activity]. This is a blank template, not a completed assessment, and it must not invent hazards, controls, legal requirements or workplace facts. Use clear British English and make it suitable for editing in a word processor or spreadsheet. Include sections for: - assessment title, location, activity and date - assessor, responsible manager and review date - people who may be harmed, including employees, contractors, visitors and members of the public where relevant - hazard and possible harm - existing controls - initial risk rating, with clearly defined likelihood and severity scales and a transparent scoring method - further controls or actions required - action owner and target date - residual risk rating after the further controls - completion status, approval and sign-off Add concise instructions beside each field. Include a table layout and a short explanation of how to use it. State that the scoring method is a planning aid and does not replace competent assessment or compliance with current UK health and safety requirements. Add a final checklist covering consultation with affected workers, implementation of controls, recording actions and setting a review trigger. Do not claim that the template is legally sufficient for every workplace. Where information is missing, mark it as [TO COMPLETE] rather than guessing.
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
What it gets wrong
- AI cannot see the workplace, observe unsafe behaviour or identify hazards that are absent from your description.
- AI cannot decide whether a proposed control is practical, adequate or properly implemented in your specific workplace.
- AI cannot replace consultation with affected workers or the judgement of a competent health and safety person.
- AI does not carry the employer's responsibility if the assessment misses a serious risk.
- A polished risk matrix can create false confidence when the underlying hazard information is incomplete.
What caps this at PARTLY: stakes of error, judgement under ambiguity and legal accountability.
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.
| Axis | Score (0–2) |
|---|---|
| Output | 2 |
| Inputs | 1 |
| Verification | 1 |
| Liability | 1 |
| Effort delta | 2 |
| Total | 7 / 10 |
The methodology and its thresholds are published in full.
FAQ
- Can ChatGPT create a risk assessment template?
- Yes, it can create the structure and wording for a blank UK workplace risk assessment template. It cannot inspect your workplace or decide whether the hazards and controls are complete, so a competent person must adapt and approve it.
- Is an AI-generated risk assessment template legally compliant?
- Not automatically. The template may support your process, but legal and workplace safety responsibilities remain with the employer or responsible manager, and the finished assessment must reflect the actual risks and controls.
- What should a workplace risk assessment template include?
- Include the activity and location, people who may be harmed, hazards and possible harm, existing controls, risk ratings, further actions, owners, deadlines, residual risk, review arrangements and approval. It should also explain its scoring method and make clear that missing information must be completed rather than guessed.
- Can I use AI to complete a workplace risk assessment?
- You can use AI to organise information and draft the document, but do not treat its completed assessment as approval. Walk through the work with affected staff and have a competent health and safety colleague check the hazards, controls and ratings before use.
Nearby answers
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.