DeNexus launches agentic AI platform for industrial cyber insurance underwriting
DeNexus has released DeRISK UWA Agentic, an AI platform designed to automate industrial cyber insurance underwriting. The platform completes underwriting assessments-from application review to actuarial output-in minutes, a process that typically requires specialist staff and weeks of manual work.
The system deploys five coordinated AI agents that independently evaluate IT and operational technology environments, identify cross-domain risks, and generate binding conditions with deadlines. Every finding traces back to its source document and model version, creating an auditable record for underwriters.
What the platform produces
DeRISK UWA Agentic outputs expected loss, maximum foreseeable loss, loss exceedance curves, premium indications, structured insurance programs, and mandatory binding conditions. The platform ingests PDFs, Excel files, CSVs, Word documents, and JSON formats, and is pre-trained on standard cyber insurance questionnaires and specialized operational technology supplements.
José María Seara, founder and CEO of DeNexus, said the platform addresses a gap in the market. "The insurance industry has tools for IT risk. It does not have tools for OT risk - not at the depth or speed that underwriting demands," he said.
How it evaluates operational technology risk
The platform incorporates real-time threat intelligence from a proprietary database updated with MITRE ATT&CK for ICS techniques. Dual IT and OT specialist agents score environments in parallel against industry standards and industrial-specific frameworks.
A feature called "Sound of Silence" treats unanswered OT-specific questions as structural risk indicators rather than missing data. This approach flags potential visibility or governance gaps that underwriters might otherwise overlook.
The actuarial engine draws from more than 300 real industrial OT deployments conducted since 2019. The system is modular-insurers can replace the embedded actuarial model with their own if needed.
Report structure and assessment framework
DeRISK UWA Agentic organizes assessments into 12 sections across four analytical layers. Reports begin with a "verdict-first triage" that presents the underwriting decision and top three binding conditions before the full body text.
The framework evaluates OT cybersecurity maturity against NIST Cybersecurity Framework and IEC 62443 standards, measuring performance across Identify, Protect, Detect, Respond, and Recover functions. It models loss scenarios across optimistic, likely, and pessimistic conditions, breaking down costs for breach response, regulatory fines, operational disruption, and business interruption.
The assessment includes regulatory exposure under GDPR, CCPA, HIPAA, PCI DSS, and state breach notification laws. Recommendations guide policy limits, retentions, estimated premiums, and prioritized mitigation measures such as incident response planning and SIEM deployment.
Market positioning and adoption
Boston Consulting Group's Agentic AI Maturity Framework for Insurance (March 2026) positions most carriers at Horizon 0-1, reflecting rule-based automation or single-agent tools. DeRISK UWA Agentic operates at Horizon 2-3 on day one, with orchestrated specialist agents handling end-to-end workflows. BCG's research recommends that specialist insurers pursue buy-and-integrate strategies for agentic AI rather than building from scratch.
The platform is available in three tiers: Standard for triage, Professional for peer review, and Enterprise for binding and API integration. It meets regulatory requirements across US/NAIC, UK/FCA, Lloyd's, and EU/EIOPA jurisdictions.
For insurance professionals looking to understand how AI agents and automation apply to underwriting workflows, or exploring broader applications of AI for insurance, this platform demonstrates how agentic systems handle specialized domain knowledge at scale.
Your membership also unlocks:
