As of 13 August 2026, AI can find unused subscriptions in your business expenses.

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 business bank and card transactions as CSV or spreadsheet files, including dates, descriptions, amounts, currencies and account or card identifiers where available.
   2. Open the export and remove unrelated personal transactions, while keeping the original file unchanged as a reference.
   3. Gather supplier invoices, contract or renewal emails, payment account details and a list of staff or teams using each service.
   4. Paste the prompt into an AI data-analysis tool and attach the cleaned transaction file or files, stating the date range and which account each file covers.
   5. Open the AI's recurring-payment table and compare every flagged transaction with the original export to confirm the dates, amounts and supplier descriptions.
   6. For each item marked needs checking or unclear, sign in to the supplier account or open the latest invoice and confirm the contract owner, renewal date, cancellation terms and actual business use.
   7. Ask the relevant budget owner or staff member to confirm whether the service is still needed, then record the decision and evidence before sending any cancellation request.

   Prompt

   I run a UK business and want to identify potentially unused subscriptions in my business expenses. I will attach a CSV or spreadsheet export of business bank or card transactions covering [DATE RANGE]. Analyse the file without inventing or estimating any data.
   
   First, identify likely recurring subscription payments by grouping transactions by supplier, description and similar amounts. Distinguish subscriptions from one-off purchases, ordinary suppliers, taxes, payroll and transfers where the data allows. For each likely subscription, report the supplier, transaction description, payment dates, payment frequency, amounts, currency, most recent payment date and the evidence for treating it as recurring.
   
   Then create a review list with these columns: supplier, likely service, evidence of recurring billing, last payment, recent payment pattern, possible reason to investigate, information I need to confirm, and confidence. Do not call anything unused unless the transaction data proves that, which it normally cannot. Mark each item as either likely active, needs checking or unclear.
   
   Do not decide what I should cancel, give tax or accounting advice, or infer business use from the supplier name alone. Flag duplicate-looking charges and payments whose amount or frequency changed. State any missing columns, date gaps, duplicate rows or limitations in the file. Keep the result concise and show the transaction rows supporting each flagged subscription. Use pounds where the file uses pounds, and do not convert currencies.
   
   After the analysis, give me a short checklist for verifying each item against the supplier account, invoice, contract, renewal terms and actual staff use before I cancel anything.

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