Radiant Logic Inc., a platform that provides identity visibility and intelligence, today announced it’s extending its services to agentic artificial intelligence to help companies control and govern this new type of autonomous software.

As autonomous AI systems proliferate, including AI agents, and begin to join employees within corporate networks, these systems are being given deep access to sensitive enterprise data, tools, privileges and context.

Radiant Logic argues that this is opening up a whole new surface that is largely invisible to most companies when it comes to what can happen when something goes wrong: AI agents don’t walk in the front door, they don’t wear a badge, have no faces, don’t gather at company meetings and can’t be brought in for debriefings.

The company’s platform provides a framework to observe agent behavior in real time irrespective of where it runs – in the cloud, on-premises or in hybrid environments – or what large language model or platform it runs on. It audits the authority, inventory and continuously scores risk and calculates per-agent risk as security posture and context change.

This is important because employees treat AI agents not just as tools that they use like they would another piece of software like email, a word processor or search, but as another person. Agents themselves use these tools with their own privileges, access rules, policies and identity controls that outlive their purposes.

The proposition of control changes when an AI agent touches many tools and data on the network. That becomes a problem when the privileges and access outlive the necessity and later a malicious party discovers it attached to the agent and exploits that vulnerability.

“In the traditional identity environment, enterprises had dozens of user directories and no unified view, and we built the platform that solved it,” said Chief Executive Dr. John Pritchard. “Today, these environments have a handful of fragmented agentic registries and the same problem.”

As more agents are being deployed across corporate environments, this problem is exploding.

Radiant Logic said every hyperscaler is already building native agentic registries, but none federates across the others. This means that enterprises spreading their resources across multiple clouds and running agentic resources locally or in a hybrid will find themselves split apart and unable to coordinate and “see” what’s happening. They have no unified view of what’s happening or what each different agent can reach.

With this release, the company is providing complete agent inventory for every AI agent across the enterprise with an authoritative record. It includes configuration, approved LLMs and accessible resources, with visualization for agent-to-agent relationships, agent-to-data dependencies, and links that show which agents work with which human employees.

The platform also validates and configures the hygiene and state of operating agents. This means that if any AI system goes out of date or stale, it can be appropriately spun down and its privileges updated, revoked or temporarily disabled when it’s idle or offline.

That’s especially important should the role, configuration, privilege, ownership or behavior of an agent change moment to moment. With real-time risk scores and alerting, security and information technologies teams can stay on top of agents before they go haywire or get exploited – through accidental internal mistakes or external malicious influence.

“Every CISO I talk to is being asked to govern AI agents now, and none of them can afford to deploy a second identity stack to do it,” said Chief Product Officer Sebastien Faivre. “We built this as a native extension of the Identity Data Platform. The same federation architecture already trusted by a third of the Fortune 100 now extends to agents across every major framework. — one authoritative record across humans, nonhuman identities and agents.”

Photo: 105 0415 8790 Still Life Black and Sherlock Holmes/photopin