The Rosetta stone of CPS: Claroty’s AI-powered library

2026-05-13 · Source: Databricks · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cybersecurity & Data Privacy, Data Science & Analytics · Depth: Advanced, long

Summary

Claroty has launched its AI-Powered CPS Library, a novel mapping engine designed to resolve the "identity crisis" in Cyber-Physical Systems (CPS) by accurately identifying industrial and healthcare hardware. A report by Claroty's Team82 found that 88% of CPS assets lack exact product codes, and 76% use codes inconsistent with vendor records, complicating vulnerability management. The library employs a hybrid architecture combining classic Entity Resolution (ER) methods with Generative AI, leveraging the Databricks Data Intelligence Platform. This system automates the identification of devices like Rockwell Automation's 1769-L36ERMS/B, linking internal codes to commercial names, specific parts, firmware versions, and relevant CVEs, a process previously requiring extensive manual effort. It has improved vulnerability attribution accuracy by 25% and provided new security recommendations for 56% of analyzed devices.

Key takeaway

For CTOs overseeing operational technology (OT) security, your teams should evaluate Claroty's AI-Powered CPS Library to gain deterministic visibility into your industrial and healthcare assets. This solution can significantly reduce manual effort in vulnerability management and improve the accuracy of security recommendations, thereby enhancing your organization's overall cyber-physical resilience and compliance posture.

Key insights

A hybrid AI system can resolve complex industrial asset identification challenges for enhanced cybersecurity.

Principles

• Deterministic traceability requires multi-agent AI.
• Domain-specific embeddings improve accuracy.
• Human-in-the-loop refines AI models.

Method

The system uses an orchestrated multi-agent AI, including NLP and Reasoning Agents, on a Databricks Lakehouse architecture. It processes diverse data, applies statistical inference, and integrates human feedback to achieve high-fidelity entity resolution.

In practice

• Automate vulnerability attribution for CPS assets.
• Implement LLM-as-a-Judge for continuous evaluation.
• Use domain-specific RAG for specialized data.

Topics

• Cyber-Physical Systems
• AI-Powered CPS Library
• Entity Resolution
• Databricks Data Intelligence Platform
• Multi-Agent AI Systems

Best for: CTO, AI Security Engineer, AI Engineer, MLOps Engineer

Related on AIssential

• How Digital Workers Are Changing Industrial Performance - with Somya Kapoor of IFS Loops — Agentic systems enable intelligent automation for complex industrial tasks, overcoming prior AI limitations.
• With $1 Cyberattacks on the Rise, Durable Defenses Pay Off — Generative AI accelerates both cyberattacks and defenses, necessitating a shift to foundational security rather than reactive patching.
• Introducing the Red Agent POV Series — AI-powered offensive security agents can autonomously discover complex, logic-driven vulnerabilities at scale, surpassing traditional methods.
• Governing Security in the Age of Infinite Signal – From Discovery to Control — AI's accelerated vulnerability discovery mandates a critical shift from detection to robust control and governance to manage escalating systemic risk.
• What Breaks First in Enterprise AI Systems — Prioritizing AI velocity over robust governance and architecture creates significant security and operational risks for enterprises.
• How dangerous is Anthropic’s Mythos AI? | Bruce Schneier — AI models are rapidly advancing in finding vulnerabilities across software and complex rule-based systems.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Databricks.