FICO, a global analytics software leader, has been awarded 10 new patents by the U.
S. Patent and Trademark Office and Canadian Intellectual Property Office, enhancing its capabilities in Responsible AI, bias detection, fraud prevention, and data privacy across its market-leading AI solutions.
These newly granted patents reinforce FICO’s position in Responsible AI innovation and applied intelligence technology. The company’s patent portfolio now includes 231 active U.
S. and foreign patents, with an additional 79 patent applications filed and pending. These innovations directly enhance core FICO solutions such as FICO® Falcon® Fraud Manager and FICO® Platform, which are used in billions of enterprise decisions globally.
The expansion of FICO’s patent portfolio highlights its commitment to advancing Responsible AI and excellence in transaction analytics. These patented technologies are designed to drive profitability, customer satisfaction, customer protection, and growth in various sectors, including financial services, telecommunications, healthcare, retail, transportation, and supply chain.
Dr. Scott Zoldi, chief analytics officer at FICO, stated, “Our patent portfolio continues to grow, and it reflects our commitment to innovation and developing technological advancements for our clients that are not just powerful, but transparent and trustworthy. We are helping clients solve problems responsibly in identifying why an AI model made a specific prediction or in detecting subtle biases across complex datasets. We’re not just advancing AI technology—we’re making Responsible AI the new competitive standard.”
The 10 new patents cover diverse aspects of AI and analytics technology:
One patent describes a “Method for Real-Time Enhancement of a Predictive Algorithm by a Novel Measurement of Concept Drift Using Algorithmically-Generated Features,” which involves a real-time monitoring algorithm to identify systematic changes in machine learning model behaviors. This technology supports FICO’s Responsible AI strategy and FICO® Falcon® Fraud Manager.
Another patent, “False Positive Reduction in Abnormality Detection System Models,” aims to reduce the fraud false-positive rate for abnormal transactions at previously used merchants or locations. This is integrated into FICO® Falcon® Fraud Manager and FICO® Fraud Predictor with Merchant Profiles.
The patent titled “Supervised Machine Learning-Based Modeling of Sensitivities to Potential Disruption” focuses on developing and utilizing machine learning to analyze historical data for model generation. It quantifies the sensitivity of predictions regarding an entity’s expected performance to future potential disruptions and is core to the FICO® Resilience Index.
“Training Artificial Neural Networks with Constraints” addresses technical improvements in training machine learning models to enhance explainability and ensure conformance with specific business requirements. These improvements are applied in FICO® Falcon® Fraud Manager, FICO® Fraud Predictor with Merchant Profiles, FICO® Open Banking Analytics, and FICO® Transaction Scores.
A breakthrough method for explaining machine learning model scores is detailed in the patent “Attributing Reasons to Predictive Model Scores with Local Mutual Information.” Integrated into FICO® Falcon® Fraud Manager and FICO® Fraud Predictor with Merchant Profiles, this technology helps financial institutions understand factors driving fraud alerts, improving transparency and regulatory compliance.
The patent “Segmentation Using Zero Value Features in Machine Learning” tackles data sparsity in production environments, improving model accuracy and reducing false positive rates by treating zero values as predictive information. This is implemented in FICO® Platform.
“System and Method for Linearizing Messages from Data Sources” optimizes the processing of data streams through tasks like unbundling, normalization, and validation. This innovation powers FICO® Decision Management Platform Streaming, enabling real-time decision-making for millions of transactions.
“Rule Based Automation” enhances software verification outcomes through improved Behavior-Driven Development testing methodologies, expanding test coverage and minimizing maintenance for complex decision systems.
The patent “Overly Optimistic Data Patterns and Learned Adversarial Latent Features” strengthens AI models against adversarial attacks by anticipating and mitigating vulnerabilities. This technology has been incorporated into FICO® Falcon® Fraud Manager and FICO® Fraud Predictor with Merchant Profiles to enhance security.
Finally, “Relationship Retrieval of a Partitioned List of Records” covers methods for efficiently identifying and retrieving relationships in partitioned records, supporting the capabilities of FICO® Identity Resolution Engine. This enhances entity resolution and relationship discovery in large datasets.
FICO (NYSE: FICO), founded in 1956, is a pioneer in predictive analytics and data science. The company holds over 200 U.
S. and foreign patents on technologies that enhance profitability, customer satisfaction, and growth. Its solutions are utilized in over 80 countries, protecting billions of payment cards from fraud, improving financial inclusion, and increasing supply chain resiliency. The FICO® Score, used by 90% of top U.
S. lenders, is a standard measure of consumer credit risk in the U.
S. and is available in over 40 other countries.
