Detecting more fraud at pre-authorization while increasing approvals

Powerful market-ready AI

Transaction Fraud Monitoring is a fraud decisioning solution providing risk scores to acquirers, payment service providers and payment facilitators at the pre-authorization stage of the payments flow.



Trained on Mastercard’s global network intelligence

Models make decisions in near real time (10 ms on-premise; 100-120 ms cloud) without needing to access external data during a transaction.

How does transaction fraud monitoring work?

Market-ready AI requires a minimal lift. Start by integrating your data flow via the cloud-based API.

Send as few as 30 data elements to initialize the model.

Risk scores are returned via the intuitive interface, including customizable Rules Management, Case Management and Business Insights modules.

How can you benefit from Transaction Fraud Monitoring?

Immediate deployment

Start seeing ROI from day one.

Increase profits

Identify fraud earlier and approve more transactions.

Harness trust

Share in the power of Mastercard’s trusted data practices using responsible, proven, transparent AI. 

Frequently asked questions

What is global network intelligence?

Mastercard’s decades of experience innovating and managing transactions provide us with a vast range of high-level transaction intelligence derived from anonymized and aggregated data. Our models are trained on this intelligence to learn patterns and anomalies from anywhere in the world Mastercard does business. Models are built following Privacy by Design standards and meet compliance locally and globally. 

How do you identify fraud at pre-authorization?

Our Transaction Fraud Monitoring solution intercepts at pre-authorization, identifying fraud before it hits other security defenses. Our AI models’ protection and accuracy are high because they’re pre-trained on the most sophisticated fraud attempts worldwide. They don’t have to refer to network data during transactions to score events, resulting in faster response times. 

What’s the difference between Mastercard’s Decision Intelligence and Transaction Fraud Monitoring?

Decision Intelligence (DI) is optimized for issuers at the authorization stage of transactions. Transaction Fraud Monitoring (TFM) is optimized for acquirers, providing risk scores on data elements at the preauthorization stage (point of sale), adding an additional layer of defense to any network transaction. 

Both solutions operate behind the scenes, analyzing transaction data in real time and flagging potential fraud for further investigation by financial institutions or payment processors. DI and TFM are available for any card brand or payments network through Access. 


What makes market-ready Transaction Fraud Monitoring easy to implement?

Because we’ve trained the model with the most difficult-to-catch fraud intelligence, further training is not required. With just a minimal amount of your data – as few as 30 data elements – to initialize the model, it’s ready to identify anomalous patterns and begin scoring.  

What kind of results can I expect?

Global transaction intelligence brings knowledge from the broader market to deliver fewer false positives and higher fraud detection rates. One global acquirer increased fraud detection 2-3 times and increased approvals by 7.4 percent.  

Can I create a custom model for Transaction Fraud Monitoring as opposed to using a market-ready solution?

Yes, you can. Using AI Express, Mastercard’s proven six-step process, we will quickly build a complex custom AI solution using Brighterion AI technology and prepare you for model deployment in a matter of weeks. 

Our team will meet with yours to discuss your business challenge. The Mastercard team comprises data scientists skilled at designing custom AI models, domain experts experienced in using AI for transaction fraud detection, and developers who build models geared to solving your business’s challenge. One of our skilled project managers will keep the team on track and liaise with your team lead.  


Experience the power of Brighterion AI