Claroty releases AI-powered device library technology

OT cybersecurity company Claroty has launched its AI-powered CPS Library across its Claroty xDome and Claroty Continuous Threat Detection (CTD) solutions.

Claroty says security teams have long struggled to accurately assess the scope and remediation of cyber risk associated with the assets in their environment, lacking a centralised standard repository that aids in uniquely identifying assets across networks. A Claroty Team82 research report, ‘Resolving the CPS Identity Crisis’, found that 88% of CPS assets currently do not transmit an exact product code, and 76% transmit product names that differ from the vendor’s official record. This leaves security teams struggling to ascertain whether they have a complete picture of the assets in their environments and can be left with a partial correlation of vulnerabilities to individual assets, creating blind spots, prolonged exposure to attacks, and incomplete remediation.

Claroty says the CPS Library is the industry’s only asset catalogue that reaches greater depths of visibility and enables precise and deterministic traceability for vulnerability attribution. To build the AI-powered device repository, Claroty partnered with automation vendors and medical device manufacturers, including Rockwell Automation and Schneider Electric, to ensure enhanced visibility and accuracy in tracking asset specifications.

“Resilience starts at visibility, and organisations require precise and deterministic traceability for every connected device that could potentially become exposed as an open door to attackers,” said Yaniv Vardi, CEO of Claroty. “The CPS Library is revolutionising the way we reduce risk by leveraging AI-driven techniques combined with our industry expertise that will drive the whole industry forward in how we improve accurate risk and exposure information.”

Claroty says the CPS Library has AI integrated at all stages, allowing for advanced asset identification, precise vulnerability mapping and automated insight generation. LLMs and statistical inference modelling are used to ingest fragmented data about assets and vendors, intelligently modelling, categorising and correlating this information into a single, actionable source of truth that significantly lessens the burden of security teams tracking critical CPS assets across complex environments.

The CPS Library is a central component to the next stage of Claroty’s AI strategy. The CPS Library enhances organisations’ ability to detect assets and gain deep insights. Additional AI capabilities include:

• The Claroty MCP Server: Customer teams can put their CPS security data to work by automating device queries, speeding up incident response, and expanding the ability to leverage this data by all relevant teams using their preferred generative AI tool.
• AI for asset identification: Statistical inference modelling, human–machine teaming, and a regionalised large language model retrieval augmented generation (LLM-RAG), work together to provide asset discovery and attribute information.
   

“We’re excited to deepen our relationship with Claroty in supporting its new CPS Library,” said Tony Baker, Vice President & Chief Product Security Officer at Rockwell Automation. “Cybersecurity is becoming even more complex in an increasingly interconnected world and is intertwined with smart manufacturing priorities that demand precise device identification. The granularity of device identification honed by the Claroty CPS asset library empowers organisations to better protect their industrial automation assets from evolving threats.”

Claroty’s CPS Library represents a major step forward for the industry,” said Jay Abdallah, President of Cybersecurity Solutions at Schneider Electric. “For too long, security teams have faced fragmented approaches to identifying and managing CPS assets, leaving gaps in risk visibility. This standardised catalogue brings clarity and consistency to an incredibly diverse asset landscape, enabling organisations to make faster, more confident decisions to protect critical operations.”

Image credit: iStock.com/Traitov