Elisity Joins Open Secure AI Alliance to Help Advance Open Source Tools and Standards for AI Safety and Cyber Defense

Company brings a collaborative ecosystem approach to industry coalition

SAN JOSE, Calif., Oct. 1, 2026 /PRNewswire/ -- Elisity, the identity-based microsegmentation company, today announced it has joined the Open Secure AI Alliance. The Alliance, led by the Linux Foundation, brings organizations together to develop and share open tools, techniques, and technologies that safeguard software and AI agents. Elisity will contribute its perspective on connecting identity and risk context with least-privilege policies, containing agentic threats, and addressing unsanctioned AI use across diverse technology environments.

AI agents can act across applications, infrastructure, and services from many providers. Securing that activity requires identity systems, threat detection, application permissions, and segmentation controls to work together. The Alliance supports collaboration across the model, harness, and runtime, as well as the broader software and security ecosystem.

“Enterprises need a consistent way to secure AI agents as they move across technologies from different providers,” said James Winebrenner, CEO of Elisity. “We have built Elisity around collaboration: bringing together the context our customers’ security tools provide and turning it into enforceable segmentation policy. We’re joining the Open Secure AI Alliance to bring that experience to the development of open tools and standards for agent identity, least privilege, and containment.”

Bringing ecosystem experience to a shared challenge

Elisity’s collaborative approach is reflected in its existing integrations with platforms including CrowdStrike Falcon, Microsoft Active Directory, Armis, Claroty, and ServiceNow CMDB. IdentityGraph correlates identity, asset, and risk context from these complementary sources to inform segmentation policies enforced through supported network infrastructure. This experience informs Elisity’s perspective on the information security systems need to exchange so that enterprises can make consistent decisions about what AI agents should be allowed to reach.

Through the Alliance, Elisity intends to work with fellow participants on open approaches to three enterprise security priorities:

• Least privilege for sanctioned agents. Establish common ways to describe an agent’s identity, owner, and authorized scope so that security controls can apply appropriate limits across platforms. Approved agents should be treated as internal actors whose access must be justified by their work and reassessed as that work or its risk changes.
• Containment of agentic threats and attacks. Develop shared approaches to communicating risk and coordinating containment across security systems. Enforceable boundaries and interoperable controls are essential to limiting lateral movement and the reach of threats that operate at machine speed.
• Identification and control of unsanctioned AI. Identify improvements to how systems share evidence of AI use outside approved deployments, including employee-installed agents and newly connected AI services. Common approaches should help organizations distinguish observed activity from indicators of possible use and connect that evidence to appropriate policy decisions.

“Healthcare organizations need to know who owns an agent, what access its work requires, and how to stop it if something goes wrong,” said Jason Elrod, CISO at MultiCare Health System and executive adviser to Elisity. “Those requirements span clinical applications, infrastructure, and security tools. Shared standards can help us apply them consistently as AI becomes part of patient care and hospital operations.”

“A new AI tool can enter through an employee’s laptop or a service the organization already uses,” said Mike Searight, former CIO of the City of Waco, Texas, and an adviser to Elisity. “Public-sector teams need their technology providers to work together so they can understand that access and limit it. Open approaches give us a better chance of doing that with the infrastructure and staff we already have.”

Elisity sees participation in the Alliance as an extension of its commitment to helping enterprises put their existing security investments to work together. The company’s vision for AI-agent security depends on collaboration across the ecosystem, with shared standards connecting identity, permissions, and enforceable boundaries as adoption grows.

Learn more about the Open Secure AI Alliance and Elisity.

Frequently Asked Questions About Elisity and the Open Secure AI Alliance

What is the Open Secure AI Alliance?

The Open Secure AI Alliance is an industry coalition, led by the Linux Foundation, that brings organizations together to develop and share open tools, techniques, and technologies that safeguard software and AI agents. The Alliance supports collaboration across the model, harness, and runtime, as well as the broader software and security ecosystem.

Why did Elisity join the Open Secure AI Alliance?

Elisity joined the Open Secure AI Alliance to bring its experience turning the context from customers’ existing security tools into enforceable segmentation policy to the development of open tools and standards for agent identity, least privilege, and containment. AI agents act across applications, infrastructure, and services from many providers, so securing them requires identity systems, threat detection, application permissions, and segmentation controls to work together.

What will Elisity work on through the Open Secure AI Alliance?

Through the Open Secure AI Alliance, Elisity intends to work with fellow participants on open approaches to three enterprise security priorities: least privilege for sanctioned agents, containment of agentic threats and attacks, and identification and control of unsanctioned AI.

What does least privilege mean for AI agents?

Least privilege for AI agents means describing each agent’s identity, owner, and authorized scope in a common way, so security controls can apply appropriate limits across platforms. Elisity’s position is that approved agents should be treated as internal actors whose access must be justified by their work and reassessed as that work or its risk changes.

What is unsanctioned AI, and why is it hard to control?

Unsanctioned AI is AI use outside approved deployments, including employee-installed agents and newly connected AI services. It is hard to control because evidence of that use is spread across many systems. Elisity supports common approaches that help organizations distinguish observed activity from indicators of possible use and connect that evidence to appropriate policy decisions.

How does Elisity connect identity and risk context to segmentation policy?

Elisity IdentityGraph correlates identity, asset, and risk context from complementary sources, including CrowdStrike Falcon, Microsoft Active Directory, Armis, Claroty, and ServiceNow CMDB, to inform segmentation policies enforced through supported network infrastructure. Elisity enforces least-privilege access without agents, new hardware, or network redesign.

About Elisity

Elisity helps healthcare systems, manufacturers, and critical infrastructure operators stop lateral movement with identity-based microsegmentation. Using identity and risk context for users, devices, workloads, and AI agents, the platform discovers assets, simulates policies, and enforces least-privilege access through existing network infrastructure—without agents, new hardware, or network redesign. Customers include GSK, Main Line Health, MultiCare Health System, and Shaw Industries. Learn more at www.elisity.com.

Media Contact:
Danielle Ostrovsky
Hi-Touch PR
Ostrovsky@Hi-TouchPR.com