en.Wedoany.com Reported - Identity management platform Lumos has announced the launch of Identity Agent Force, a set of AI agents designed to continuously manage access permissions for people, machines, and AI agents within enterprises. This agent suite is integrated into the new Agent Hub, with the initial lineup including Access Review Agent, Access Request Agent, Role Mining Agent, Entitlement Analyst, NHI Owner Hunter, and Agent Ownership Finder, with more agents planned for release later this year.

Attackers are leveraging AI agents to launch attacks. For instance, using attack models like Mythos, a single attacker can accomplish the work of a hundred people, conducting phishing, credential theft, and lateral movement at speeds surpassing human teams. Meanwhile, defenders need to protect more entities than ever before, including humans, non-human identities (such as service accounts and API keys), and the emerging AI agent workforce, all of which require unified management and protection.

Security-focused identity management teams cannot rely on traditional, human-centric identity tools to address the current situation. Ticketing, reviews, and one-time requests have already overwhelmed teams, and security gaps are widening every quarter.

If attackers have agents, defenders need agents too. Identity Agent Force provides enterprises with a means to counter AI with AI: an integrated set of agents that automate manual tasks at machine speed and on a large scale in the background. Human efforts will shift toward using direction, strategy, and best practices to build systems, rather than processing tickets.

Andrej Safundzic, CEO of Lumos, stated that AI is reshaping enterprises at an unprecedented pace, and traditional identity management faces a dilemma: either move slowly and become obsolete, or act quickly and expose the enterprise to risk. Attackers will not wait, and defenders need the same advantage to fight back and win.

Unlike traditional identity platforms that rely on periodic, manual workflows (such as requests, approvals, reviews, and certifications), Identity Agent Force changes the operational model. Each agent runs continuously in the background, making access decisions for all identities, escalating only anomalies that require human judgment. When a team needs more coverage, another agent can be deployed from the Agent Hub.

Each agent operates based on two foundations: first, real-time mapping of every identity and every permission, covering every human user, non-human identity, and AI agent, down to permissions within each application; second, memory of the company's actual operational context, such as who has what permissions, which approvals require legal department involvement, the permission boundaries for sales engineers, and the customer data contractors can access. Lumos builds and maintains this context, enabling each new agent to understand the business from day one of deployment.

Specific agent functions include: Access Review Agent runs end-to-end access reviews, self-certifying secure access and presenting only portions requiring human decisions; Access Request Agent grants exactly the needed access and automatically revokes it on time, with permanent permissions gradually reduced with each request; Role Mining Agent learns actual team permission usage patterns and drafts least-privilege roles within seconds; Entitlement Analyst translates permission language into plain English, helping approvers accurately understand the actual scope of access granted; NHI Owner Hunter monitors every service account, API key, and token, shutting down dormant or overly broad permissions before attackers discover them; Agent Ownership Finder catalogs agents and non-human identities running in the business and assigns a human owner to each entity.

Because Lumos is built on a complete Identity Governance and Administration (IGA) foundation, these agents can go beyond one-time fixes, turning every action into a lasting strategy and improving security posture and compliance with each decision. Lumos is the first identity platform built around autonomous agents rather than manual workflows. The platform currently serves teams at companies such as Mars, Netskope, Assurant, and GitLab, helping them move faster, reduce risk, and demonstrate compliance while maintaining human control.

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com