FICO® Focused Foundation Model for Financial Services
A breakthrough in responsible, applied AI. The FICO® Focused Language Model and FICO® Focused Sequence Model are domain-specific, genAI models and deliver higher accuracy and explainability for critical business decisions.
State of AI today
FOCUS is all you need
General purpose LLMs excel at broad applications, but lack the domain-specific knowledge and precision required for critical financial service use cases. Focused models bridge this gap by combining the power of advanced AI with deep industry expertise, regulatory compliance, and the explainability that financial institutions demand.
This targeted approach augments your AI strategy with purpose-built intelligence that unlocks new value for financial institutions.
Powerful things happen with our unique, customer-focused approach
Domain-specific
with FICO® Intelligence
and Responsible AI
For language use cases and beyond
FICO® Focused Foundation Model for Financial Services has two components, each with a revolutionary model architecture that unlock a spectrum of new and powerful use cases:
Focused Language Model
Focused Language Model
The FLM addresses natural language use cases. It provides trustworthy responses with sensitive knowledge for use with both employees and customers. This approach dramatically improves regulatory compliance efforts, unleashes the power of unstructured data, and empowers you with the ability to deploy genAI for use cases where it was previously not possible.
Focused Sequence Model
Focused Sequence Model
The FSM empowers you to detect patterns in transaction data and customer behavior. It complements traditional transaction analytics with the power of our patented AI innovation and deep domain expertise in real-time payments. The flexible design means it’s not limited to fraud detection and can unlock any use case where you need to predict the next action.
What sets our foundation model apart
Purpose-built for financial services, FICO FFM is trained on curated financial data derived from a patent-pending technique. This results in sharper, more reliable results than general-purpose models.
Expert-defined reference points containing specialized domain vocabulary that guide FICO FFM’s outputs for specific tasks.
Patent-pending Trust Scores assess reliability of outputs by checking alignment against knowledge anchors and coverage statistics on training data. Trust Scores prevent hallucinations, supports compliance, and guide human-in-the-loop decisions.
"FLMs are essentially similar to SLMs (Small Language Models) and these are transforming how GenAI is used in financial risk management and compliance by providing highly accurate, domain-specific insights and reducing misinformation. Built on curated data and Responsible AI principles, these models are becoming essential tools for institutions that require precision, transparency, and scalable trust,”
Megha KumarRelated resources
How Trust Scores Can Help Businesses Rely On GenAI Outputs
A risk-based approach to measuring GenAI accuracy through focused language models and knowledge anchors.
2025 State of Responsible AI in Financial Services: Unlocking Business Value at Scale
How AI and technology leaders in financial services are unlocking business value through Responsible AI.
GenAI in the Past, Present, and Future
Navigating the evolution of AI with responsible deployment strategies and proven governance frameworks.
Worried about Gen AI Hallucinations? Focused Language Models are an Imaginative and Proven Solution
Why financial institutions are turning to specialized AI models to mitigate regulatory and operational risks.