Vercel has introduced Enterprise Apps and Agents, a set of capabilities that place internal applications and AI agents behind a company’s identity provider by default. The centerpiece, Vercel Passport, authenticates access to those apps and agents against an organization’s existing employee identity system rather than leaving them reachable without an enterprise login.
Passport is built to plug into the identity stack a company already runs. It supports Okta, Microsoft Entra, Auth0, and other OpenID Connect-compatible providers, and applies authentication against employee identity, private-by-default settings, full auditability, and centralized administrative policy across app and agent deployments. The effect is to make an internally deployed app or agent subject to the same access controls as any other corporate system, instead of being a separate surface with its own ad hoc access.
The agent piece is where identity gets harder. As organizations deploy AI agents that act on internal systems, those agents need to authenticate and be governed like any other actor, and Vercel pairs Passport with a connection layer that gives agents short-lived credentials for systems such as Slack, GitHub, Snowflake, and Salesforce rather than static, long-lived keys. Short-lived credentials limit the blast radius if an agent or its credentials are compromised.
The launch reflects a broader move to bring AI agents under enterprise identity governance rather than treat them as unmanaged software. Vendors have been building identity layers specifically for autonomous agents, as with Ping Identity’s Identity for AI to secure and govern the autonomous agent economy and Anonybit’s decentralized platform for AI agent identity security. Vercel’s approach routes agents through the same identity provider and policy controls a company already uses for its workforce.
Vercel Passport is available in public beta, with the connection layer for agent credentials in beta alongside it.
Sources: Vercel
–
By the ID Tech Editorial Team
