What's new in Immuta

As AI agents increasingly become the primary way users interact with data platforms, traditional access control models are reaching a breaking point. When any user can ask an agent to perform complex tasks, managing individual database permissions becomes a daunting and often dangerous task.

At Immuta, we believe that access can no longer be the gatekeeper to data; policy must be.

⚡ We are excited to announce our new solution for Agentic Data Access, designed to secure your data in this new architectural era. Key highlights include:

• Agents as first-class identities: Treat AI agents as distinct identities with specific access policies defined by attributes, groups, and tags.
• Zero standing privileges: Immuta vends the user’s specific data access policies to the agent only at the time they are needed. Once the job is done, the vended role is deleted. There is no need for that user to have standing privileges in the data platform.
• Seamless integration: Even if a user (like someone in HR) doesn’t have a specific account in your data platform (e.g., Snowflake), they can still safely interact with data through an agent leveraging that user's vended access role.
• Continuous control: We are bridging the gap between authentication and authorization, ensuring that policy follows the user through the agent to the actual data access point.

👉 Learn more in our blog posts:

• The Agentic Breaking Point: Why AI Demands Centralized Authorization.
• Solving the Agentic Breaking Point: Why AI Demands a New Blueprint for Data Access

🔎 Read about governing AI agents with Immuta in the documentation.

Agentic data access is available in private preview. Contact your Immuta representative for details.