Dun & Bradstreet launches agentic AI for enterprise workflows

Dun & Bradstreet announced an agentic AI platform that combines its commercial datasets with autonomous AI agents to automate business research, risk assessment and customer outreach for enterprise clients.

The platform accepts high-level goals-such as vetting suppliers, scoring counterparty risk or generating lists of leads-and executes multi-step workflows without continuous human intervention. Models can query D&B datasets and produce structured outputs aimed at credit, procurement and sales teams.

D&B integrates its business identifiers, firmographic, financial and corporate linkage data with natural language models and rule-based controls. Outputs include narrative summaries and machine-readable files that can feed downstream systems.

Agents can access live data feeds, refresh outputs when underlying data changes, and connect to enterprise systems via APIs and common authentication methods so organizations can link the platform to customer relationship management, enterprise resource planning and risk systems.

Governance features include configurable approval gates, activity logs and restrictions on which data sources an agent may use. Those controls are designed to preserve human oversight where required and provide traceability for decisions such as onboarding a customer or adjusting a credit limit.

Dun & Bradstreet positioned the product for large commercial customers that manage complex supplier networks, credit portfolios or sales pipelines. The company will offer professional services and onboarding support to help clients map agent workflows to internal processes and compliance requirements.

In testing scenarios presented by D&B, agents reduced the time needed to compile supplier due-diligence packets and automated routine credit reviews. The company noted that complex decisions would continue to require human judgment and formal approvals.

The launch follows a broader trend of software vendors combining large language models with proprietary datasets to automate workflows. Agentic AI refers to systems that plan and execute multi-step tasks on behalf of a user. Businesses are seeking controls on data access, model behavior and audit trails when those systems handle sensitive commercial information.