Moody’s CEO: AI’s Trust Deficit Remains Unresolved Despite Model Advancements

(SeaPRwire) –   Weekly news cycles are frequently centered on the debut of the latest AI models. Recently, Meta introduced Muse Spark, its inaugural model from a restructured AI unit, which internal tests suggest is on par with top industry competitors across various functions.

However, the continuous release of new models highlights a growing trend: as the market becomes saturated, these models are increasingly viewed as commodities. This raises a critical question for companies looking to scale AI: what will serve as the primary differentiator?

The solution is found in a single concept—trust.

In the long run, the specific model being utilized will be less important than the “connected intelligence” that informs it. Connected intelligence refers to high-quality data gathered from various structured sources, allowing an AI to process a comprehensive set of information rather than relying on a partial view.

To use an analogy: AI models are like vehicles that are constantly being upgraded, while data and intelligence serve as the navigation system. A basic GPS using an old map might eventually get you to a destination, but it lacks reliability and speed.

Perhaps it will work.

But “perhaps” is insufficient for high-stakes environments, particularly in the financial sector. We are dealing with critical decisions regarding loans, insurance affordability, and protection against financial crimes. These AI models require a definitive source of truth; otherwise, we risk poor results and the further erosion of public trust in global institutions.

NVIDIA CEO Jensen Huang recently emphasized that “structured data is the ground truth of AI.” He pointed out a reality the industry has been slow to accept: a sophisticated model is only as good as the trusted data behind it, and not all data meets that standard.

Data must be meticulously organized and calibrated to reflect real-world conditions. This process involves more than just web scraping. Organizations that combine advanced models with this type of connected intelligence will foster trust and ensure their AI-driven decisions are justifiable to regulators, boards, and shareholders.

The pitfalls of a weak data foundation are already evident. MIT reports that 95% of AI pilot programs fail to produce significant results, largely due to inadequate data. Superior models cannot rectify this; in fact, they can make flawed outputs more difficult to identify and harder to correct.

Current global events—such as shifting tariffs, geopolitical changes affecting supply chains, extreme weather, and infrastructure cyberattacks—illustrate the dangers of inaccurate outputs. As noted in the World Economic Forum’s Global Risks 2026 report, risks are becoming more massive, interconnected, and fast-moving.

For the financial industry, this interconnectedness is a practical reality. In an era of “Exponential Risk,” threats are not just larger; they are linked. For instance, a weather disaster can disrupt a supply chain, subsequently affecting economic growth and credit. Generic AI paired with disjointed data cannot provide a reliable risk assessment. Conversely, connected intelligence that integrates data on climate, credit, and compliance offers a much more trustworthy path forward.

By unifying various data sources, companies can gain a more precise and actionable understanding of risk than siloed data allows. Businesses that integrate third-party data with their own will be better positioned to make rapid, defensible decisions.

While AI models have become significantly more capable over the last few years, the focus must now shift to the quality of the intelligence powering them. This is not just a task for engineers; it is a priority for anyone aiming to leverage AI effectively. Organizations must challenge their data teams and vendors to ensure their intelligence is connected, reliable, and proven against real-world outcomes.

The implications extend beyond profit and growth to the very foundation of institutional trust that markets depend on. Moody’s was established over a century ago on the belief that markets thrive on transparent and independent analysis. That principle remains vital today; AI does not change it, but it does increase the stakes of failing to uphold it.