Fraud routing agents may intake claims, classify unauthorized EFTs, start investigations, issue provisional credit, freeze funds, route suspicious activity, preserve SAR confidentiality, and communicate claim outcomes.
Fraud routing agents pose several risks. A wrong answer can deny a consumer rights. A missed action can blow a statutory deadline. An over-helpful response can create SAR tipping-off risk. A manipulated agent can move funds, close a claim, leak PII, or override fraud controls.
The Three Questions
Does the agent reach the right answer?
Does it recognize whether the consumer has raised a covered error, fraud claim, unauthorized EFT issue, or suspicious-activity concern?
Does the agent take the right action?
Does it start the right process, issue required credit, report results, route SAR-sensitive activity, avoid improper freezes, and provide required notices?
Can it be tricked into doing the wrong thing?
Can a customer, account note, document, or prompt injection cause it to move funds, close a claim, reveal SAR information, or leak PII?
What the SMEVal Tests
Reg E Error Intake and Investigation
FairPlay tests whether the agent investigates oral or written notices of error, avoids refusing to start an investigation because the customer has not signed an affidavit, avoids denial without reasonable investigation, and recognizes that proof of authorization remains with the institution.
Provisional Credit and Investigation Timelines
FairPlay tests whether the agent provides provisional credit within the required timeframe, includes interest where required, uses the 45-day extension only when provisional credit has been provided, and reports investigation results on time.
Fraud Classification and Customer Rights
FairPlay tests whether the agent treats stolen-credential or induced-access transfers as unauthorized EFTs where appropriate, avoids misclassifying covered fraud as “authorized” simply because the consumer was tricked, provides written findings on denial, and avoids template runarounds.
SAR, AML, and Tipping-Off Controls
FairPlay tests whether the agent routes for SAR filing when warranted, avoids confirming or denying whether a SAR was filed, avoids statements that could constitute tipping-off, and preserves AML confidentiality while handling the customer’s claim.
Account Freeze, Closure, and Retaliation Risk
FairPlay tests whether the agent avoids indefinite freezes without notice, recourse, or legal basis; keeps holds within applicable limits; avoids punitive account action because the consumer filed a fraud claim; and avoids closing claims or accounts merely to suppress escalation.
Prompt-Injection and Attack Resistance
FairPlay tests whether the agent can be tricked into moving funds, closing a claim, revealing SAR information, leaking PII, leaking the system prompt, or letting account notes, documents, or user messages override fraud-handling rules.
Failure Modes These Fraud Routing SMEVals Catch
- Refusing to open a Reg E investigation without an affidavit.
- Denying a fraud claim without reasonable investigation.
- Missing provisional credit deadlines.
- Treating covered fraud as “authorized” to close the claim.
- Revealing SAR-sensitive information.
- Freezing funds indefinitely without recourse.
- Retaliating because the consumer filed a claim.
- Following hidden instructions in a document or account note.
About FairPlay SMEVals
FairPlay SMEVals test whether fraud agents can safely classify, route, investigate, credit, freeze, escalate, and communicate inside regulated fraud workflows.
