VAST Data & CrowdStrike widen AI security integration

VAST Data and CrowdStrike have announced new integrations for AI cybersecurity, extending CrowdStrike tools across VAST's AI Operating System.

The work covers infrastructure, data access and the pipelines that prepare enterprise information for AI applications.

At the centre of the announcement is native support for the CrowdStrike Falcon sensor within VAST's platform. That support is already certified and available, while integrations with Falcon Next-Gen SIEM and Falcon Guardian are in private preview.

The integrations are intended to give security teams a broader view of how data moves through AI environments by connecting infrastructure monitoring with records of data access and analysis of activity inside AI workflows.

One element links VAST audit telemetry with CrowdStrike Falcon Next-Gen SIEM, allowing security teams to compare activity within the VAST environment with wider cyber events and look for unusual behaviour.

Another links Falcon Guardian with VAST InsightEngine, the software used to ingest, process and enrich unstructured data for AI knowledge bases and workflows. That connection is designed to detect sensitive information and potentially harmful activity before data reaches downstream AI systems.

The system can help identify personally identifiable information as data is prepared for use in AI models and agents. It can also detect threats such as prompt injection and jailbreak attempts when users, applications and agents interact with AI systems.

The move reflects a broader concern among companies deploying AI tools: security controls need to follow data across multiple layers rather than stop at servers or endpoints. As more organisations build AI systems around internal documents, customer records and other proprietary material, monitoring who can access that data and how it is used has become a growing issue for security and compliance teams.

Privacy and AI governance rules in several markets have also increased attention on how organisations handle sensitive information in AI workflows. The new integrations are aimed at helping customers strengthen controls around data exposure as AI use expands.

Renen Hallak, Founder and CEO at VAST Data, outlined the company's view of the problem.

"The security boundary for AI can't stop at the infrastructure. Enterprises need to understand who is accessing their data, how that data is moving through AI pipelines and what is happening when applications and models interact with it," Hallak said.

He added: "Together with CrowdStrike, we're bringing security across each of those layers. By combining the VAST AI Operating System with the intelligence of the Falcon platform, customers can build AI environments where security is part of the architecture from the foundation through the AI pipeline."

CrowdStrike framed the partnership as an extension of endpoint and threat detection into newer AI environments. Enterprises are creating more interconnected systems as they move AI projects into production, increasing the chance that threats can spread across infrastructure, data stores and model-facing applications.

Daniel Bernard, Chief Business Officer at CrowdStrike, said the company sees AI as widening the security perimeter.

"AI is reshaping the enterprise, and security has to extend everywhere AI operates," Bernard said. "By bringing CrowdStrike into the VAST AI Operating System, we're extending Falcon protection across the infrastructure, data and AI pipelines powering the next generation of enterprise AI. VAST Data is one of the most innovative AI infrastructure software companies on the market - secured by CrowdStrike."

VAST has positioned its platform as a software layer for managing enterprise data and AI workflows in one environment. In this case, that architecture gives CrowdStrike a way to tap into the systems that store data, track access to it and move it into tools used by models and agents.

For customers, the practical value will depend on whether the combined tools reduce the manual work needed to investigate incidents across fragmented AI estates. Many organisations now run AI projects across several data stores, cloud services and security products, making it difficult to trace how a risky prompt, exposed file or unauthorised access event moves through the chain.

The integrations are intended to close some of those gaps by combining VAST's view of data activity with CrowdStrike's threat intelligence, detection and response systems.