FAQs for the FraudClassifierSM Model
What is the FraudClassifier model?
- The FraudClassifier model is a fraud classification structure that can be used to classify fraud involving payments.
- It offers a holistic picture to help organizations better understand the magnitude of fraud involving payments.
- The key advantage of the model is the ability to classify fraud independently of payment type, payment channel or other payment characteristics.
- The model can be downloaded and used at no cost. There is no obligation or requirement to use the model; use is entirely voluntary. The model and definitions are available for download by registering for the model.
How does the FraudClassifier model work?
- The FraudClassifier model is built upon who initiated the payment and the actions around the event involving the payment. This was done in order to facilitate a clear, intuitive and objective approach to classifying fraud.
- To further promote consistent application of the model across the industry, supporting definitions for key terms are included with the model.
- The model guides users through a series of three questions in order to help classify fraud according to the model construct.
- The first question is “Who initiated the payment?” and focuses on whether the payment was initiated by an authorized or an unauthorized party.
- The second question examines how the fraud was conducted.
- The final question focuses on the specific tactic(s) and approach(s) used to carry out the fraud.
- Dependent on how much information is available at the time of the initial fraud classification, you may answer one, two or all three questions. The model is built to be flexible, allowing classification additions or modifications as new information becomes available.
Why should an organization consider adopting the FraudClassifier model?
- To facilitate consistent fraud information and tracking. The FraudClassifier model can be applied across an organization to help ensure greater internal consistency, more robust information and better fraud tracking.
- To improve customer education. The FraudClassifier model may help an organization better understand fraud to more effectively educate its customers on current fraud methods and how to protect themselves from being victimized.
- To better understand fraud across payment types and fraud methods. The FraudClassifier model was designed to help classify fraud for multiple payment types and fraud methods. Such insights can improve fraud response strategies.
- To speak the same language about fraud. If adopted across the payments industry, the FraudClassifier model can facilitate a common fraud language and help us work together to better identify and fight fraud, fostering a safer payment system for all.
How can the FraudClassifier model offer an organization a more holistic view of fraud?
- The FraudClassifier model allows an organization to obtain a holistic picture of both authorized and unauthorized party fraud. It can be used to classify fraud initiated by either an authorized party or unauthorized party – two types of fraud that demand different fraud management responses.
- Fraud initiated by an authorized party is often overlooked and underreported. By using the model, which includes both, an organization can understand the “big picture of fraud.”
Is use of the FraudClassifier model mandatory?
- No. Adoption of the FraudClassifier model is completely voluntary, and its release does not create any regulatory requirements, supervisory expectations or mandates in any way. Voluntary sharing and use of the model are encouraged industrywide. Learn more by reviewing the industry adoption roadmap and the disclaimer at the bottom of this page.
Who developed the FraudClassifier model, and for what purpose?
- The Federal Reserve established and led the Fraud Definitions Work Group to help address the industrywide challenge of inconsistent classifications for fraud involving ACH, wire or check payments.
- The work group met in 2019 and 2020, and brought together relevant expertise and a range of experience from across the payments industry. A full list of members can be found on the Fraud Definitions Work Group webpage.
Can the FraudClassifier model be used to classify fraud beyond ACH, wire or check?
- The FraudClassifier model was designed with the flexibility to extend beyond ACH, wire and check payments, and to incorporate other fraud types.
Did the Fraud Definitions Work Group (FDWG) get input from payments industry stakeholders on the FraudClassifier model before finalizing? How was it tested to ensure it offers appropriate coverage of fraud in the industry?
- In October 2019, a draft of the FraudClassifier model was shared with members of the Fraud Definitions Community Interest Group (CIG), a subset of the FedPayments Improvement Community members that have indicated their interest in this topic, with an accompanying testing exercise to gauge the overall intuitiveness of the model.
- The CIG is comprised of 1,300 industry stakeholders and subject matter experts who signed up to follow FDWG developments and provide input on its deliverables.
- CIG members were asked to classify payments fraud scenarios using the model.
- The survey responses provided valuable insights into areas for further refinement and affirmed that the model is intuitive for payments stakeholders.
What is the industry adoption roadmap proposed by the Fraud Definitions Work Group?
- The industry adoption roadmap is a strategy and potential steps to encourage voluntary industrywide use of the FraudClassifier model to achieve a more consistent dialogue on fraud, both within and across organizations. It recommends two actionable paths for advancing industry implementation – adoption of the model within organizations, and integration of the model into industry fraud studies.
- Voluntary industrywide adoption of the model could serve as an important step toward improving the consistency of fraud classification and providing the ability to speak the same language about fraud.
Does the FraudClassifier model involve any data storage?
- The FraudClassifier model was designed for fraud classification purposes and does not inherently collect, track or store data. Individual organizations decide independently how to use and implement the model, including where and what data to store, as well as who has access to the data.
- The Federal Reserve does not collect, track or store any model data from model adopters.
Is there any cost to obtain or use the FraudClassifier model?
- There is no cost to obtain or use the FraudClassifier model, definitions or supporting classification tool.
How can interested parties learn more about the FraudClassifier model and start using the model? Where can questions about the model, or ideas to improve it, be directed?
- Visit the FraudClassifier model webpage to learn more and register to access educational resources and support tools for the model as a member of the FedPayments Improvement Community, such as a PDF of the model (including supporting definitions for key terms) and a fraud classification tool that incorporates the model.
- Use this form to contact the Federal Reserve with questions or comments.
Why is registration needed to gain access to the FraudClassifier model and definitions?
- The Federal Reserve is requesting registration to help gather industry feedback about overall usability and enhancement opportunities. Additionally, the inclusion of other payment and fraud types will be evaluated, allowing the Federal Reserve to update industry stakeholders who signed up.
Are there examples of how various fraud scenarios may be classified according to the FraudClassifier model?
Sharing and use of the FraudClassifier model throughout the industry is encouraged; any adoption of the FraudClassifier model is voluntary at the discretion of each individual entity. The FraudClassifier model is not intended to result in mandates or regulations, and does not give any legal status, rights or responsibilities, nor is the FraudClassifier model intended to define or imply liabilities for fraud loss or create legal definitions, regulatory or reporting requirements. Absent written consent, the FraudClassifier model may not be used in a manner that suggests the Federal Reserve endorses a third-party product or service.
©2020 Federal Reserve Banks. Materials are not to be used without consent.