WorkOS builds identity and authorization layers for AI agents
WorkOS is developing a suite of products to address the emerging market for agent identity and authorization. The company argues that current human-centric infrastructure is insufficient for the scale of future AI agent interactions.
WorkOS is developing a suite of products to address the emerging 'Auth4Agents' market, focusing on data access, authorization, and identity. The company has launched WorkOS Pipes, an integration product that securely connects AI agents to customer data sources such as Google Drive, Salesforce, HubSpot, and Slack. WorkOS is currently building an authorization layer to define granular permissions and restrict agent behavior to prevent unauthorized actions. Additionally, the company is working on an identity layer for agents, drawing inspiration from Microsoft Entra agent ID and expansions to OpenID Connect.
WorkOS CEO Michael Greenwich noted that the scale of agent identity will far exceed human identity, with trillions of agents expected to operate in the future compared to 7-8 billion humans. This shift requires new approaches to logging, permissioning, and authentication systems that can handle the high volume and automated nature of agent interactions. Greenwich argues that the current identity and authentication infrastructure designed for humans is insufficient for the scale and speed of AI agents, necessitating a new paradigm of agent-specific identity and permission management.