Angemeldet als:
Standard Working GroupS
Federated Machine Learning Standards (active)
The Federated Machine Learning Working Group is developing data exchange standards to enable financial institutions to share partial detection patterns, creating a more comprehensive and effective detection model. By advancing holistic AML/CTF approaches, the group aims to reduce reliance on rigid regulatory coverage requirements, lowering detection costs while improving accuracy. To complement these technical standards, a governance framework will ensure robust and responsible implementation.
Status: Active - Design phase
Members: Charles Camp, Jean Voigt
Product & Data Description Standards (active)
The Product and Data Standards Working Group is establishing unified input data standards for customer, transaction, and reference data used in financial crime detection. In addition, standards to return true positive reports from competent authorities are prepared. Standardization will enhance industry collaboration, enable actionable pattern analysis—even without full machine learning integration—and serve as a foundation for Federated Machine Learning standards.
Status: Active - Design phase
Members: Jean Voigt, Anonymous
Name List Standards (proposed)
The Name List Standards Working Group is streamlining sanction list data distribution to eliminate duplication and reduce misinformation. The group is also developing list removal protocols and governance frameworks, ensuring data subjects have a mechanism to contest erroneous inclusions. These standards will support vendors and regulators in efficiently managing sanction and AML screening processes.
Status: Proposed - ongoing evaluation
Typology Data Examples (proposed)
The Typology Data Examples Working Group is developing a repository of standardized data examples for training and testing purposes, aligned with customer, transaction, and list data standards. These examples enhance collaboration among financial institutions, enabling more effective discussions and actionable insights. Additionally, they provide investigators with tangible representations of emerging patterns, improving their ability to detect financial crime more efficiently compared to abstract textual case studies and typology descriptions.
Status: Proposed - ongoing evaluation