Akeyless, a leader in identity security, announced Agentic Runtime Authority, a new runtime control capability for AI agents, along with Agentic Identity Intelligence, extending its AI Agent Identity Security solution. Trusted by Fortune 500 organizations across retail, financial services, healthcare, and pharmaceuticals, and securing over 220 billion machine identity interactions, Akeyless brings proven enterprise scale security to the next generation of autonomous systems.
As AI agents evolve from passive assistants to autonomous actors, they are no longer just reading data; they are modifying systems, executing workflows, and making decisions across production environments. Traditional approaches based on static credentials, OAuth tokens, and role-based access controls were not designed for this shift and lack the ability to evaluate intent, context, and real-time behavior.
With this launch, Akeyless enables organizations to discover every agent, authorize by intent, and control every action in real time, including instantly blocking unsafe behavior. This closes a critical security gap as AI agents move beyond SaaS integrations to interact with internal systems, databases, and infrastructure.
Also Read: ScaleOps Raises $130M Series C at Over $800M Valuation to Lead the Future of Autonomous Cloud and AI Infrastructure
Runtime Authority for AI Agents
Akeyless Agentic Runtime Authority introduces a new approach by enforcing security at the moment of action, not just at access. It provides Intent-aware authorization where dynamic access adapts in real-time to each request. It also enforces policy in real-time, ensuring only authorized actions are executed. As agents act in milliseconds, this enforcement provides immediate control, blocking unsafe actions within autonomous systems as they occur.
At runtime, an AI agent may require access to multiple systems across clouds and environments. In addition to intent-aware access provisioning, Akeyless enforces Zero Standing Privileges (ZSP) and enables secure, Just-In-Time (JIT) access across both modern cloud services and traditional enterprise systems, including on-premises applications, internal databases, and infrastructure. This enables organizations to operate AI agents securely across environments. Throughout execution, sessions are continuously monitored with the ability to revoke or terminate instantly. With full forensic traceability and auditability, every action is tied back to the originating prompt, providing a powerful governance mechanism.
Extending to Identity Intelligence and Visibility
While Runtime Authority enforces control in real time, Agentic Identity Intelligence provides continuous visibility across AI environments, delivering a system of record for agent identities, access, and data interactions. In addition, the system tracks data lineage across agent-driven workflows and detects risks such as excess privileges, orphaned credentials, and unauthorized access. It also provides audit-ready evidence linking agent activity to relevant policies and data.
Together, these capabilities give organizations full visibility into how AI agents access systems and data, complementing runtime control with continuous intelligence.
“AI agents are no longer just generating answers, they are taking real actions across critical systems,” said Oded Hareven, CEO of Akeyless. “Enterprises do not just need visibility, they need control. For years, Akeyless has secured billions of machine identity interactions for some of the world’s largest organizations, eliminating standing privileges and enabling just-in-time access across complex environments. With Agentic Runtime Authority, we are extending that proven foundation to AI agents, bringing intent aware authorization, real-time enforcement, and the ability to stop unsafe actions before they happen. This is what it takes to safely operationalize AI at scale.”
SOURCE: PRNewswire