Verisk elected Pradip Patiath to its board of directors on May 20, 2026. He is a McKinsey senior partner who has co-led the firm's North American digital insurance and consumer banking practices for a decade, and before that was president and chief operating officer of CCC Information Services, one of Verisk's direct competitors. He replaces a retiring audit committee chair, which is the part of the announcement that says what the board thinks it needs.

Key Takeaways

  • 34.5 billion records over 50 years is Verisk's statistical database, 8.2 billion commercial lines and 21.5 billion personal lines, and the asset the AI distribution strategy is built to move.
  • Seven AI modules shipped in Q1 2026 against a full-year target of 25, on revenue of $783 million with subscriptions at 84% of the total.
  • 44% of insurance executives say governance or compliance problems contributed to AI projects failing or underperforming, and only 24% are very confident of passing an independent AI governance review within 90 days.
  • Four to five years is the average length of Verisk's client contracts, which is the term over which a shift in who controls the interface would show up in pricing leverage.

What the Seat Replaces

Patiath joined McKinsey in 1996 and has been a senior partner since June 2011, advising insurers, banks, wealth managers, private equity and payments firms on strategy, AI and digital transformation, M&A and turnarounds across North America, Europe, South America and Asia.

The CCC years matter more for what Verisk is doing now. CCC Intelligent Solutions competes with Verisk directly, and Patiath ran it as president and chief operating officer while it scaled into one of the sector's leading digital platforms. He has built and sold enterprise analytics to carriers from inside a vendor, not only advised on it.

The succession is the clearest signal. Kathleen Hogenson had served since 2016 across the Audit, Executive and Risk committees and chaired Audit. Trading a chair with that profile for a digital transformation and AI strategist is a deliberate change in the board's competency mix, and Verisk's proxy describes the Risk Committee as strengthening oversight of the risks and opportunities in the evolving AI landscape.

From Selling a Destination to Selling a Layer

Verisk's product cadence in Q1 2026 was module by module: seven AI releases against a target of 25 for the year, augmented underwriting generating more than 20 follow-up meetings with carrier prospects, aerial imagery up more than 30% over two years, digital media forensics onboarding a sixth top-10 carrier.

The MCP connector launch on May 5 is a different kind of move. Putting ISO Indications and XactRestore inside a foundation model through an open protocol stops selling analytics as a place the user goes and starts distributing them as a layer inside a tool the user already has.

That reaches actuarial work directly. ISO loss costs, catastrophe models and the filing data that support a rate indication start arriving through an AI interface rather than a portal, which changes what has to be documented about them. A number pulled into a filing exhibit through a model-mediated call carries a different provenance question than the same number pulled from a subscription product, and the governance controls that make ISO loss costs acceptable to a state reviewer attach to the delivery mechanism as much as to the data.

It also puts the customer relationship one layer further away. Verisk's contracts run approximately four to five years by chief executive Lee Shavel's account, which is long enough for control of the interface to matter when they are renegotiated. If the analytics arrive inside someone else's assistant, the experience and eventually the pricing leverage move with it.

The three competitors are solving the same problem from other directions and none of them is choosing this one. Guidewire embedded its first agentic capability, the Underwriting Assistant, inside its own core platform in the December 2025 Olos release, making switching cost the lever. CCC sits in the auto physical damage and bodily injury claims workflow across more than 125 insurers and 15,000 repair facilities, where AI solutions now run roughly 10% of revenue and grow at 3.5 times the company rate. EXL pairs analytics with managed services, and data and AI revenue reached 60% of its mix. Verisk is the one betting on distribution through a third party's model.

The Buyer's Compliance Gate Is Where the Layer Stops

Verisk disclosed the constraint itself on the Q1 call. Shavel said the company is having to spend more time working through AI governance and compliance requirements in contract negotiations with carriers, which lengthens procurement and slows the conversion of an AI pipeline into recognised revenue.

The buyer-side numbers explain why. 44% of insurance executives report governance or compliance problems contributing to AI projects that failed or underperformed, and 24% are very confident they could pass an independent AI governance review within 90 days. Fully integrated adopters are almost four times more likely than pilots to report AI-driven revenue growth, 58% against 15%, which means most of the addressable base is on the wrong side of a gate that is not about the product.

Board enthusiasm does not close it. Three in four boards have approved major AI investments, 52% have set clear AI governance expectations and 54% have integrated AI risk into ongoing board or committee oversight. The approval came first and the oversight has not caught up.

The regulatory side reaches the vendor rather than stopping at the carrier. Twenty-five states and jurisdictions have adopted the NAIC Model Bulletin on AI governance, and the NAIC ran a nine-state pilot of its AI System Evaluation Tool from January 2026 through September. When a vendor's analytics feed a rate filing, the examiner's questions about model governance follow the data back up the chain, and an open-protocol delivery layer is a longer chain than a portal was.

That is the bind the governance friction describes. The layer strategy is a bet on being everywhere the carrier's AI already is, made into a market where the thing slowing purchases down is the carrier's inability to demonstrate control over the AI it already has.

Further Reading on actuary.info