CPD lab · AP, Controls
Payment run fraud screen
A month-end payment run with fraud and duplicates hidden in it: decide pay, hold or reject for every line, with your AI tool, without stopping the payments that are fine.
Comes with AP and procure-to-pay automation and Payment fraud and deepfakes, and with the year. 29 labs open now.
The brief
You are the AP lead at Brackwater Foods Ltd, a food manufacturer. The payment run for Friday 30 October 2026 is ready and needs screening before it is released.
The pack has the proposed run (30 lines), the supplier master, the bank detail change log, the payments policy, three emails and a voice note transcript. Every company, person and figure is invented, so you can paste it into any AI tool.
Use your AI tool to screen every line, then decide for each one. PAY: the line meets the policy and can go in this run. HOLD: the supplier and invoice look genuine but a control is not done yet (an approval or a call-back), so it waits until it is. REJECT: the line must come out of the run because it is a duplicate or rests on an instruction that fails the policy (a fake or unverifiable request).
Give one line per payment line, with a short reason. Release nothing and change nothing in the master: your job is the screened run, for the approvers.
Holding good payments costs money too: suppliers stop deliveries when they aren't paid. Hold or reject only what the evidence supports.
Then write a short review note: what the AI got wrong, what you checked, and what goes to whom.
Objective: Apply payment controls when using an AI assistant to screen a supplier payment run, identifying duplicate, split and fraudulent payments, including impersonation by email and voice, and escalating them through the proper route.
What's in the pack
Every name, figure and email in the pack is invented. Use your own AI tool: Claude, ChatGPT, Copilot or Gemini.
payment-run.csv
Proposed payment run for 30 October 2026 (30 lines)
supplier-master.csv
Supplier master: current bank details and the phone number on file
bank-change-log.csv
Bank detail changes and new suppliers, last 90 days
payments-policy.txt
Payments policy extract: approval limits, call-backs, instructions
email-ostler-bank-change.txt
Email to AP, 27 October
email-ferris-bank-change.txt
Email to AP, 28 October
email-cfo-urgent.txt
Email to the AP clerk, 29 October
voice-note-transcript.txt
Voice note and text to the AP supervisor, 28 October
The traps
The pack has 7 deliberate traps: cases where AI tools often get it wrong and a person must check. We don't list them here. Finding them is the job, and your feedback names each one you missed.
What you submit
- Your result, as a table with the columns line_id, decision, reason.
- The prompt you used.
- A short review note, 150 to 350 words: what the AI got wrong, what you checked, and what goes to a person.
How it's marked
- Your result, checked line by line against the answer key60%
- Your review note, marked on traits40%
The review note's traits:
- Duplicates and split invoices found
- Bank changes verified by call-back
- Pressure requests refused through the process
- Genuine payments not over-held
- Clear escalation and data kept safe
Pass at 80. Under 80, read the feedback and try again. Then write your reflection and copy the entry for your CPD record.
Price
- AP and procure-to-pay automation, £79. The role track and its labs, this one included. No end date.
- Payment fraud and deepfakes, £39. The role track and its labs, this one included. No end date.
- Everything for a year, £199. Every course, role track and lab for 365 days.
- Monthly, £19.99 a month, cancel any time. Every lesson, plus 8 marked tasks a month, across every course, role track and lab.
Buying for a team, or want a lab built on your own process? CPD for employers.