Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Important
Biometric authentication and fraud detection capabilities can be enabled independently or in combination depending on your use case. The use of one doesn't require the use of the other.
Fraud detection is a voice-biometric capability that compares the voice of each caller against your organization's watchlist of fraudster voiceprints stored within your Dataverse environment and raises an alert when it finds a potential match, so you can respond before a fraudster completes a fraudulent transaction.
Important
You're responsible for configuring and using this fraud detection functionality in compliance with applicable law. You're solely responsible for any fraud-risk determinations, fraudster designations, and fraud watchlists used in connection with this functionality. Microsoft doesn't recommend, suggest, or make any such determinations or designations. Don't use the output of the fraud detection functionality, including any match or no-match results, to evaluate, determine, or inform an individual's eligibility for credit, employment, insurance, housing, or any other purpose governed by the Fair Credit Reporting Act (FCRA) or similar laws. The fraud detection functionality is intended solely for fraud prevention and authentication purposes. Microsoft doesn't provide this functionality as a consumer reporting agency, and the functionality doesn't generate or provide consumer reports. You will defend and indemnify Microsoft against any claims made by a third party related to this fraud detection functionality. By checking I agree, you attest that your organization will comply with these obligations.
How fraud detection works
Voice-biometric fraud detection follows three core steps. Together, they turn a single identified fraudulent call into ongoing protection against the same fraudster on future calls.
- Detect a fraudulent call. A fraud analyst reviews interactions in Fraud Management Cases and detects a call that was made by a fraudster.
- Create a fraudster voiceprint. From that call, the fraud analyst creates a fraudster voiceprint, saves it within your Dataverse environment and adds it to your organization's watchlist.
- Generate fraud alerts. For incoming calls where fraud detection is enabled, each caller is checked against the watchlist. When a caller's voice matches a fraudster on your organization's watchlist, the system generates an alert for customer service representatives, fraud analysts, or both.
Key concepts
Understanding a few core terms helps you work with fraud detection and the related management tasks.
Watchlist
A watchlist is a list or group of fraudster voiceprints that fraud detection uses to monitor and match incoming callers. Your organization starts with an empty default watchlist, and your fraud analysts can create additional watchlists to organize voiceprints in a way that fits how different types of calls are distributed within your organization.
Fraudster voiceprint
A fraudster voiceprint is a secure mathematical representation of the distinctive characteristics of your fraudster's speech. It's a mathematical model of your fraudster's speech rather than the audio itself. Fraudster voiceprints are stored within your Dataverse environment. Fraud detection compares the voice on an incoming call against the voiceprints in your organization's watchlist and generates a fraud alert when it detects a similar voice.
Your organization's fraud analysts create fraudster voiceprints through a dedicated Fraud Management Cases UI.
Fraud severity and fraud match score
Each case carries a Fraud severity and a Fraudster score:
- Fraud severity shows the potential risk of the interaction. If fraud severity is green or low, your customer is authenticated. If it's red or medium, your customer failed authentication. If it's purple or high, your customer matched the fraudster on the watch list.
- Fraudster score shows how closely a caller's voice matches a fraudster voiceprint on your watchlist. A higher score means a stronger match.
Roles and separation of duties
Fraud detection supports separation of duties so that the work of investigating fraud is divided among distinct roles. It introduces a Fraud analyst role, which investigates calls and manages voiceprints, and a Fraud supervisor role, which oversees the fraud analysts' work, deletes fraudster voiceprints, and curates watchlists to improve the overall fraud detection performance.