Dun & Bradstreet brings verified risk data into Claude for underwriting

Dun & Bradstreet moved its Commercial Graph into Claude on July 8, giving carriers and brokers a way to pull verified business information and risk intelligence without bouncing between separate underwriting tools. The company said the integration can compress research-heavy commercial underwriting work that usually takes days or weeks into minutes, with the biggest payoff in submission intake, eligibility checks, and early referral decisions.

The setup ties D&B’s Commercial Graph and predictive analytics to Claude through a Model Context Protocol server. In practical terms, D&B is trying to turn Claude into an underwriting front end that can handle know-your-business checks, entity resolution, screening, ownership insights, financial risk checks, and onboarding in one session. D&B said a carrier intake team could clear a new submission in seconds by verifying identity, ownership structure, broker of record, producer licensure, regulatory requirements, duplicate submissions, and financial risk before handing over a decision-ready profile to an underwriter.

That architecture matters because it is aimed at the exact swivel-chair work that slows commercial P&C. Instead of asking underwriters to chase company records, cross-check risk flags, and stitch together data from multiple systems, D&B wants the answer to live inside the conversation. The company said the results are intended to be explainable, auditable, and consistent, which is the non-negotiable test in regulated insurance workflows where every referral and declination needs a clear trail. Anthropic has already pushed Claude in that direction, saying on May 5 that it released ten ready-to-run finance agent templates and that its finance connectors give Claude governed, real-time access to provider data.

D&B has been widening that same data layer across other AI interfaces. On June 16, it said the Commercial Graph would be brought into Claude, ChatGPT/Codex, and Microsoft 365 Copilot, and that verified business information is essential for agentic workflows in research, onboarding, compliance, finance, and risk decisioning. The company also keeps grounding the pitch in the structure of its own identity graph: the Commercial Graph is anchored by the D-U-N-S Number, introduced in 1963, and supported by more than 100 billion verifications, tests, and checks each month.

The timing is no accident. D&B’s May AI Momentum Survey found that 97 percent of organizations had active AI initiatives, but only five percent said their data was adequately ready to support them. The same survey said 56 percent planned to increase AI investment over the next 12 months and 60 percent reported at least some measurable ROI. D&B also said on June 18 that its agentic compliance and third-party risk capabilities could cut processing times by 70 percent to 90 percent. At Insurtech Insights USA 2026, held June 3 to 4 at New York’s Javits Center, more than 6,000 attendees and more than 400 speakers filled the conversation around the same problem: how to get trusted data into underwriting workflows fast enough to change quote-to-bind performance.