Radiant Logic Extends Identity Visibility to AI Agents

Radiant Logic, an identity-security vendor, said on June 4 that it is extending its Identity Visibility and Intelligence Platform (IVIP) to the "agentic enterprise," adding continuous, real-time risk scoring for AI agents, as reported by SiliconANGLE and in a Business Wire release. The company says IVIP now inventories AI agents across environments regardless of where they were built or run, captures each agent's configuration, approved model, and accessible resources, links it to a human owner, and maps agent-to-agent, agent-to-tool, and agent-to-data relationships.

Radiant Logic describes the update in three parts. Complete Agent Visibility builds the inventory and relationship graph. Security Posture Assessment validates configuration hygiene, flags stale or orphaned agents, checks privilege against an agent's purpose, and produces a per-agent risk score that recalculates as posture, privilege, ownership, and behavior change. Automated remediation then acts as risk rises, which the company says happens "before the blast radius compounds."

CEO Dr. John Pritchard argued that enterprises once ran dozens of fragmented user directories with no unified view and that Radiant Logic built the platform to solve that, framing today's scattered agent registries as the same problem in a new form, according to the company. Radiant Logic also states that 80% of the Fortune 500 are actively deploying AI agents, a figure presented by the company rather than an independent finding, and it positions itself as a platform-agnostic authority that works across vendors.

AI agents create persistent nonhuman identities that hold entitlements to tools and data, and their access can outlive the human who created them, a gap the release calls an "Uncontrolled Inheritance Chain." Security and identity teams increasingly have to extend inventory, entitlement review, and runtime telemetry from human users to agents. Approaches like this typically require integrations with identity providers, cloud agent registries, and security telemetry, and a platform-agnostic claim still demands engineering to normalize metadata and event semantics across providers.

This is a vendor product announcement, and the coverage reviewed is based on the company's release without third-party validation. Worth watching: named connectors for major hyperscaler agent registries, published risk-scoring methodology or schemas, references to emerging agent-identity standards, and independent customer evidence that the platform discovers cross-cloud agents and reduces real-world risk.