Ti3 is here: A bigger graph for a fast moving fraud landscape

Plaid has launched Trust Index 3 (Ti3), an advanced machine learning model for its fraud detection system, Plaid Protect, which leverages an expanded fraud graph to enhance detection capabilities across its network. Ti3 introduces a tenfold increase in the network map's size, enabling deeper real-time analysis of relationships between devices, accounts, identities, sessions, and institutions, thereby identifying coordinated fraud patterns that single platforms might miss. The model can analyze up to nine degrees of separation, detecting complex fraud activities such as device rotations and synthetic identity schemes with improved accuracy and maintaining the same false positive rate as previous models. Ti3 also offers new actionable signals and attribute categories to help fraud teams better understand and respond to emerging attack vectors, emphasizing account behavior, device activity, identity patterns, and network relationships. Enhanced device identification through Plaid Link and improved account age intelligence further bolster its capabilities, reflecting Plaid's commitment to continuously evolving its fraud prevention strategies by utilizing its vast network of real-world financial activity data.