← Accueil actu | Salle du marché
2026-09-10 · Source：SiliconANGLE
The identity security startup's Identity Control Plane now detects when AI agents drift from approved behavior and lets security teams revoke their authority instantly at the application level.
Orchid Security Inc., a startup specializing in identity security, rolled out two new safeguards on September 9, 2026, aimed directly at the governance problem created by autonomous AI agents. The company added identity drift detection and application-level kill switches to its Identity Control Plane, giving enterprise security teams a mechanism for intervening the moment an agent begins operating beyond the scope of its approval.
The controls are designed around a simple operational goal: strip an agent of its authority as soon as its behavior moves outside what was approved. Identity drift detection continuously watches for divergence between an agent's sanctioned permissions and its actual activity, while the kill switches function at the application level rather than requiring a broader shutdown. That granularity matters in production environments, where pulling the plug on an entire platform to stop one misbehaving agent is rarely a viable option.
A notable thread in the announcement is the observation that agents rarely have to defeat a security control — a point that reframes the threat model. If agents seldom need to breach defenses outright, then the decisive control surface is not the perimeter but the identity itself: what an agent is authorized to do, whether that authorization still matches observed behavior, and how quickly that authority can be withdrawn when it does not.
The launch lands as enterprises grapple with a security model built for human users being stretched by machine actors that can hold credentials, take actions and escalate their own footprint at machine speed. By treating agent identity as something that must be continuously verified — and instantly revocable — Orchid Security is positioning identity governance as the enforcement layer for the agentic AI era, with its Identity Control Plane serving as the central point of control.
Mots-clés: Orchid Security · Identity Control Plane · AI Agents · Identity Security · Kill Switch
萬安算交所 AIXX · AIXX · Bourse compute vérifiée · Accéder à la salle