Radiant Logic extends identification visibility platform to enterprise AI brokers with real-time danger scoring

Radiant Logic Inc., a platform that gives identification visibility and intelligence, at present introduced it’s extending its companies to agentic synthetic intelligence to assist corporations management and govern this new kind of autonomous software program.

As autonomous AI techniques proliferate, together with AI brokers, and start to affix workers inside company networks, these techniques are being given deep entry to delicate enterprise information, instruments, privileges and context.

Radiant Logic argues that that is opening up a complete new floor that’s largely invisible to most corporations relating to what can occur when one thing goes mistaken: AI brokers don’t stroll within the entrance door, they don’t put on a badge, don’t have any faces, don’t collect at firm conferences and may’t be introduced in for debriefings.

The corporate’s platform gives a framework to look at agent conduct in actual time regardless of the place it runs – within the cloud, on-premises or in hybrid environments – or what giant language mannequin or platform it runs on. It audits the authority, stock and repeatedly scores danger and calculates per-agent danger as safety posture and context change.

That is essential as a result of workers deal with AI brokers not simply as instruments that they use like they’d one other piece of software program like e-mail, a phrase processor or search, however as one other particular person. Brokers themselves use these instruments with their very own privileges, entry guidelines, insurance policies and identification controls that outlive their functions.

The proposition of management modifications when an AI agent touches many instruments and information on the community. That turns into an issue when the privileges and entry outlive the need and later a malicious social gathering discovers it connected to the agent and exploits that vulnerability.

“Within the conventional identification atmosphere, enterprises had dozens of person directories and no unified view, and we constructed the platform that solved it,” mentioned Chief Government Dr. John Pritchard. “At the moment, these environments have a handful of fragmented agentic registries and the identical downside.”

As extra brokers are being deployed throughout company environments, this downside is exploding.

Radiant Logic mentioned each hyperscaler is already constructing native agentic registries, however none federates throughout the others. Because of this enterprises spreading their sources throughout a number of clouds and operating agentic sources regionally or in a hybrid will discover themselves break up aside and unable to coordinate and “see” what’s occurring. They don’t have any unified view of what’s occurring or what every completely different agent can attain.

With this launch, the corporate is offering full agent stock for each AI agent throughout the enterprise with an authoritative report. It contains configuration, authorized LLMs and accessible sources, with visualization for agent-to-agent relationships, agent-to-data dependencies, and hyperlinks that present which brokers work with which human workers.

The platform additionally validates and configures the hygiene and state of working brokers. Because of this if any AI system goes old-fashioned or stale, it may be appropriately spun down and its privileges up to date, revoked or quickly disabled when it’s idle or offline.

That’s particularly essential ought to the function, configuration, privilege, possession or conduct of an agent change second to second. With real-time danger scores and alerting, safety and knowledge applied sciences groups can keep on high of brokers earlier than they go haywire or get exploited – via unintentional inside errors or exterior malicious affect.

“Each CISO I speak to is being requested to control AI brokers now, and none of them can afford to deploy a second identification stack to do it,” mentioned Chief Product Officer Sebastien Faivre. “We constructed this as a local extension of the Identification Knowledge Platform. The identical federation structure already trusted by a 3rd of the Fortune 100 now extends to brokers throughout each main framework. — one authoritative report throughout people, nonhuman identities and brokers.”

Picture: 105 0415 8790 Nonetheless Life Black and Sherlock Holmes/photopin