AI Check Kiting Detection Playbook
A bank examiner has flagged three business accounts at your institution that appear to be kiting: using float between accounts at two banks to inflate available balances. The accounts collectively processed $4.7M in check transactions in 30 days. None of the checks have cleared on normal schedules.
When to use this playbook
- Use this playbook when the decision looks like the situation above: A bank examiner has flagged three business accounts at your institution that appear to be kiting: using float between accounts at two banks to inflate available balances.
- It is a fit when you have source files in hand and need a structured, reviewable analysis — not a generic chat answer about "Check Kiting Detection".
- Do not use it as a substitute for licensed, legal, clinical, or authorized official judgment in the domain.
What you'll need
- Transaction history for all three accounts (90 days)
- Check clearing times and return records
- Interbank transfer records
- Account opening and beneficial owner documentation
- Relationship between the three account holders
Attachments: Documents (Documents)
The Prompt
You are a fraud analyst investigating check kiting across three business accounts. I am attaching: Work only from the attached source files. If a conclusion is not supported, say so. Produce: 1. Map the circular flow of funds: identify all transactions where a check deposit at Bank A funds a withdrawal at Bank B before the deposit clears. 2. Calculate the average float days being exploited and the maximum inflated balance across all three accounts on any single day. 3. Identify whether the pattern accelerated over time (a sign the kiter is approaching account closure before detection). 4. Assess the beneficial owner relationships: are these accounts controlled by the same individual or entity? 5. Tell me whether to freeze the accounts now or continue monitoring, and what documentation I need before I file a SAR. Call out where independent models are likely to disagree, and list follow-up documents a reviewer should request.
What to expect
- Circular transaction flow map
- Float inflation timeline
- Account acceleration analysis
- Beneficial owner link assessment
- SAR filing recommendation with documentation checklist
Review before you act
- Validate this output against source files before relying on it: Map the circular flow of funds: identify all transactions where a check deposit at Bank A funds a withdrawal at Bank B before the deposit clears.
- Validate this output against source files before relying on it: Calculate the average float days being exploited and the maximum inflated balance across all three accounts on any single day.
- Validate this output against source files before relying on it: Identify whether the pattern accelerated over time (a sign the kiter is approaching account closure before detection).
- Validate this output against source files before relying on it: Assess the beneficial owner relationships: are these accounts controlled by the same individual or entity?.
- Confirm every cited figure, date, counterparty, or requirement against the attached originals — models compress and can drop a qualifier.
- Treat disagreement between models as a review item, especially on classification, materiality, and recommended next action.
- Do not authorize an operational, clinical, legal, credit, or enforcement action solely because the models agree.
Why compare models on this
For Check Kiting Detection, running the same attachments across independent models is useful because the hard part is classification and completeness, not fluency. The workflow is already designed to surface circular transaction flow map; float inflation timeline; account acceleration analysis; beneficial owner link assessment. Those are comparison artifacts — they only exist if more than one model runs. Typology labels (bust-out vs. first-party vs. third-party) and ring membership often diverge across models when data is incomplete. Divergence is a reason to hold and verify, not to auto-file a SAR.
See governed multi-model AI on your own prompt
Compare GPT-5, Claude, and Gemini side by side, with human review and a decision record built in.

