As organizations adopt AI agents, expand cloud infrastructure, and connect more applications and services, identity is becoming a central part of cybersecurity. Traditional approaches to access management were designed primarily around human users, but modern environments must also account for machine identities, automated workflows, and systems that can act independently.
A new generation of founders is working to address these challenges by rethinking how digital identities are verified, governed, and protected. From identity infrastructure to AI-driven security, these leaders are developing approaches designed to help organizations maintain control as their digital environments become more complex.
Armin Ebrahimi, Co-founder and CEO of LayeredTrust
Armin Ebrahimi brings decades of experience across identity, artificial intelligence, and security to LayeredTrust, a company focused on establishing trust and accountability for AI agents operating in enterprise environments. As organizations move AI agents from experimentation into production, conventional identity and access management systems face a new challenge: authenticating an agent does not necessarily establish whether it should be allowed to perform a particular action.
LayeredTrust is addressing this distinction between access and authority. Its approach focuses on verifying what an AI agent is permitted to do, under whose authority it operates, and whether that authorization remains valid when an action takes place. Ebrahimi’s vision centers on giving enterprises stronger controls over autonomous systems without creating unnecessary friction for employees and developers.
Sundar Krish, Founder and CEO of AgntID
Sundar Krish is building AgntID around an emerging challenge in enterprise security: AI agents can dynamically select tools, interact with applications, and execute tasks in ways that traditional identity systems were not designed to manage. A technology entrepreneur with experience spanning cybersecurity, DevSecOps, and AI infrastructure, Krish previously founded sken.ai, which was acquired by Fortinet in 2021. That background informs his focus on bringing stronger security controls to increasingly autonomous systems.
AgntID is developing runtime access control for AI agents, with an emphasis on making authorization deterministic, tightly scoped, and enforced within the infrastructure where agents operate. Rather than relying solely on static permissions established in advance, the platform aims to control what agents can do as they execute tasks. This approach gives organizations a way to address the risks associated with agents accessing sensitive resources or taking actions beyond the scope of their intended responsibilities.
Pranay Yadav, Founder and CEO of Iden
Pranay Yadav founded Iden to tackle persistent problems in identity governance, particularly the manual work required to manage employee access across an organization’s applications. Many businesses rely on a combination of spreadsheets, access tickets, and custom integrations to handle onboarding, offboarding, and permission reviews. These processes can leave gaps in visibility and create unnecessary administrative work for already stretched IT and security teams.
Iden provides an identity governance platform designed to manage identities, applications, and permissions across a company’s technology stack, including applications that do not support standard automated provisioning protocols. Its tools help automate access changes, manage employee lifecycle events, and maintain visibility into permissions for both human and non-human identities. Yadav’s approach focuses on making comprehensive identity governance easier to deploy and operate, particularly for organizations that lack dedicated teams to manage complex identity infrastructure.
Dan Hollinger, Founder of INDRA Identity
Dan Hollinger founded INDRA Identity to address the growing mismatch between traditional identity systems and the rise of autonomous software. Enterprise environments increasingly depend on service accounts, API keys, automated pipelines, and AI agents, yet many identity controls were originally built around people signing into applications. Hollinger argues that these machine-driven interactions require a different approach to identity, one that accounts for the specific task being performed and the duration for which access is needed.
INDRA Identity is developing runtime governance and intent-based access controls for AI agents and other machine identities. Its model uses ephemeral credentials and narrowly scoped authorization, giving each agent or sub-agent an identity tied to its task, with access revoked when the work ends. The platform also aims to create tamper-evident records of actions, helping security teams reconstruct activity and investigate incidents. Hollinger’s vision is to bring identity controls closer to where autonomous systems actually operate, rather than relying exclusively on static credentials and periodic reviews.
William Lin, CEO of AKA Security
William Lin leads AKA Security, which evolved from AKA Identity as the company expanded its focus toward securing AI-driven development and agent workflows. Lin brings experience as a cybersecurity investor and company builder, including his work at Forgepoint Capital. His perspective on identity security has increasingly centered on how AI systems use credentials, interact with sensitive information, and execute actions within development environments.
AKA Security focuses on building security controls for teams developing with AI. Its work includes AI Traffic Control, an open-source tool designed to inspect and govern prompts and tool calls within an AI agent’s execution environment. The company also provides security engineering support and tools intended to help organizations protect credentials, control outbound connections, and establish safer configurations for AI coding agents. Lin’s broader objective is to bring security directly into the environments where AI agents operate and take action.
Identity Security for an Autonomous Enterprise
As AI agents become more capable and interconnected, identity security is expanding beyond the traditional question of who can log in. Organizations must also determine what an identity can do, whether its permissions match its purpose, and how its actions can be monitored and verified.
The work underway across identity governance, runtime authorization, and AI security reflects a broader shift in enterprise defense. Stronger controls over human and machine identities will be essential as businesses delegate more tasks to autonomous systems. The challenge ahead is to make those controls comprehensive enough to reduce risk while remaining practical for organizations adopting AI at scale.
