As of 13 August 2026, AI can analyse your employees' working hours.

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. Export the relevant time records from your rota, time-recording or payroll system as a CSV or spreadsheet, including employee identifier, date, start time, finish time, recorded hours, scheduled hours, breaks, absence and team where available.
   2. Remove unnecessary personal information, replace names with employee IDs, and check your organisation's rules before uploading employee data to an external AI service.
   3. Open a spreadsheet and record the source file name, export date, column meanings, time zone, whether hours are decimal or clock time, and how breaks and overnight shifts are represented.
   4. Paste the redacted data and the copyable prompt into an AI data-analysis tool, then ask it to produce the totals, exceptions, charts and reproducible calculations.
   5. Compare the employee and period totals in the AI response with independently calculated spreadsheet totals, and inspect every flagged row against the original export.
   6. Ask the AI to correct any discrepancy using the source rows, then save the final tables, assumptions, data-quality list and calculation steps with the original export.
   7. Send questions about pay, contractual hours, rest breaks, working-time compliance or disciplinary action to your payroll, HR or employment-law adviser rather than treating the analysis as a decision.

   Prompt

   Analyse the employee-hours data below as a reporting assistant. The data is for a UK workplace.
   
   Data:
   [PASTE A REDACTED CSV OR TABLE HERE]
   
   Column meanings:
   [DESCRIBE EACH COLUMN, INCLUDING WHETHER HOURS ARE DECIMAL HOURS OR HOURS AND MINUTES]
   
   Produce:
   1. Total recorded hours by employee and by pay period.
   2. Total recorded hours by date, team and location where those fields exist.
   3. A comparison of scheduled hours and recorded hours where both exist.
   4. Missing values, duplicate rows, impossible durations, overlapping shifts and other data-quality issues.
   5. Employees or periods with unusually high or low hours compared with their own records or the relevant team, showing the exact figures and the comparison used.
   6. Clear tables and simple charts suitable for a manager.
   7. A short list of findings, with each finding linked to the rows or calculations supporting it.
   8. The formulas or calculation steps used so I can reproduce every total in a spreadsheet.
   
   Rules:
   - Do not invent missing values or infer why a pattern occurred.
   - Treat employee names and identifiers as confidential and use anonymised labels where possible.
   - Do not decide whether the data complies with UK employment law, payroll rules or a contract. Mark those questions for a payroll, HR or employment-law professional.
   - State any assumptions before using them.
   - If the data is insufficient for a requested result, say exactly what is missing.
   - Recalculate the key totals independently and report any discrepancy.
   - Do not recommend disciplinary action or make claims about an employee's performance or conduct.

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