FICO introduced a tool designed to help companies combat artificial intelligence hallucinations.

The FICO Foundation Model for Financial Services (FICO FFM) is designed to help financial services firms derive “sustained value” from generative AI models, the company said in a news release Tuesday (Sept. 23).

“The focused foundation model represents a practitioner’s approach to gen AI in financial services, moving beyond trying to refine universal knowledge models,” FICO Chief Analytics Officer Dr. Scott Zoldi said in the release. “FICO FFM enables enterprises to use [small language models (SLMs)], built for their specific business problems, significantly helping to mitigate hallucinations, provide transparency, auditability and adaptability. The model complies with regulations through the transparency of data and a decreased risk of hallucinations through trust scores and business owner-defined knowledge anchors.”

Domain-specific models can lead to a 38% uptick in compliance adherence use cases and a more than 35% increase in world-class transaction analytic models in areas like fraud detection, the release said.

FICO FFM consists of FICO Focused Language Model for Financial Services (FICO FLM) and FICO Focused Sequence Model for Financial Services (FICO FSM), according to the release.

FICO FSM is designed to uncover “critical relationships in transaction histories” that aren’t typically found by traditional analytics systems, the release said. This helps bolster real-time detection accuracy across financial services transaction analytics, while capturing “complex inter-relationship sequences” in scenarios like payment fraud and real-time risk assessment.

Advertisement: Scroll to Continue

Meanwhile, “FLMs are essentially similar to [SLMs], and these are transforming how gen AI is used in financial risk management and compliance by providing highly accurate, domain-specific insights and reducing misinformation,” Megha Kumar, research vice president of analytics and AI analyst at IDC, said in the release. “Built on curated data and responsible AI principles, these models are becoming essential tools for institutions that require precision, transparency and scalable trust.”

The PYMNTS Intelligence report “Fighting Fraud and Finding Trust Amid Banking’s Data Deluge” found that banks, credit unions and FinTechs are rethinking data strategies around financial crime prevention as AI takes on a greater role.

Finance leaders said in the report that “data remains indispensable, but its reliability depends on balance, blending historical records with real-time signals, human oversight with machine intelligence, and institutional rivalry with shared intelligence,” PYMNTS reported Wednesday (Sept. 17).

For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.