In this rapidly evolving AI landscape, companies are racing to implement artificial intelligence solutions across their operations. However, as I discussed with Dr. Scott Zoldi, Chief Analytics Officer at FICO, in our recent episode of The Age of AI, it’s not enough to simply be fast — businesses need to ensure AI is being used responsibly and ethically, and that’s where AI trust scores and blockchain are expected to play an outsized role.

AI Trust Scores and Blockchain, the Foundation for the Golden Age of AI — Watch the full interview with Dr. Zoldi:

Building a Proper Foundation for AI Implementation

Before diving into AI initiatives, Dr. Zoldi recommends companies ask themselves four critical questions:

Why do we need AI? Not every problem requires an AI solution. Determine if traditional technologies might provide acceptable results before pursuing AI implementation.

Do we have the right expertise? The complexity of AI is mind-boggling, making it difficult for any single person to master all technologies. Companies should identify their areas of strength and determine where additional expertise might be needed.

How will we operationalize this? Understanding latencies, throughputs, computational budgets, and inference methods before beginning ensures your AI will fit within operational constraints.

What standards will we follow? Following a standardized framework allows for consistent development, monitoring, and governance of AI solutions.

I would add a fifth consideration: examining the state of your data management practices. At the end of the day, AI is simply a representation of the data that feeds it. Without proper data management, AI initiatives will struggle to deliver value.

How to Establish Trust in AI Systems

Companies build trust in AI through accountability and adherence to development standards. Dr. Zoldi outlined several components of a trust-building approach:

  • Robust AI: Ensuring data is representative and reviewed for bias.
  • Interpretable AI: Choosing appropriate machine learning types for regulated industries where understanding what the AI has learned is crucial.
  • Ethical AI: Testing learned relationships for proxies of bias.
  • Auditability: Setting clear, immutable requirements and success criteria upfront.

When these principles are followed, trust is maintained as models consistently perform well and deliver expected outcomes.

AI Trust Scores: Managing Risk in Generative AI

One of FICO’s innovative approaches involves AI trust scores, particularly for generative AI. These scores measure how well a generative AI’s answers align with “knowledge anchors” —expected answers and relevant vocabulary for specific questions.

“Trust scores help with [hallucination],” Dr. Zoldi explained. “For an organization or enterprise like our customers, they dial up or dial down the trust scores based on the type of decisions or outcomes that would come from these generative AI models.”

This risk-based approach allows organizations to operationalize generative AI technologies while limiting hallucinations and other potential issues.

Blockchain’s Crucial Role in AI Governance

In addition to AI trust scores, blockchain can also play a role in AI governance. Zoldi is very bullish on the power of AI and blockchain combined, something I’m sure we’ll be hearing much more about moving forward. Zoldi shared how at FICO, blockchain provides an immutable record of how AI models are developed, ensuring compliance with AI standards.

“What’s important about blockchain is that what we record there is an immutable record of the truth,” he noted. This technology enforces adherence to AI standards by recording requirements, testing procedures, and success criteria. If requirements aren’t met, models aren’t released.

Beyond responsibility and governance, this approach significantly improves productivity. “The blockchain is a tremendous tool for time to market,” Dr. Zoldi emphasized. By prescribing algorithms and success criteria, organizations can develop high-quality, trustworthy AI models more efficiently.

The Golden Age of AI for Financial Services

What makes this the “golden age” of AI for financial services? According to Dr. Zoldi, it’s the unprecedented combination of powerful AI technology, abundant computing resources, responsible frameworks, and specialized platforms that can operationalize these capabilities.

For financial institutions navigating their AI journeys, the possibilities are tremendous — “like a kid in a candy store,” as Dr. Zoldi put it. New frameworks and tools continue to emerge, allowing organizations to use AI more aggressively while maintaining ethical responsibility and safety.

As businesses embrace this golden age, AI trust scores and blockchain should be key considerations. Establishing trust through consistent standards, leveraging blockchain for governance, and implementing trust scores for risk management will be essential foundations for successful AI implementation.

See more of my work here:

Beyond the Copilot Button: How Agentic AI is Transforming Asset Management

Voice AI: The Revolution Transforming Customer Service Operations