Fasoo AI says AI agents need data policies beyond identity controls

Fasoo AI says enterprise security teams need data-level governance, not just identity checks, as AI agents gain more autonomy over sensitive business information. The Seoul-based company argues that policy must travel with the data to control how AI views, uses and shares files across systems.

Why it matters: - AI agents are moving beyond simple access checks and into workflows that can retrieve, summarize, generate and share sensitive business information. - Fasoo AI says organizations need controls that follow the data itself, or AI adoption can expose documents, datasets and internal workflows to unnecessary risk.

What happened: - Fasoo AI on Oct. 27 said identity-based security is not enough to govern how AI agents interact with sensitive enterprise data. - Jason Sohn, executive managing director at Fasoo AI, said each AI agent needs its own identity, but identity alone only decides whether the agent can access a system. - Sohn said data-level policies must govern how an agent views, uses and shares sensitive information once access is granted.

The details: - AI agents can now retrieve information, generate content, interact with multiple business systems and execute actions with growing autonomy. - The company said these agents can summarize documents, prepare reports, access knowledge repositories and coordinate workflows across enterprise environments. - Fasoo AI said authentication should not be the end of security decisions when an AI agent enters a business application. - Access to one application does not justify unrestricted access to every document, dataset or action involving sensitive information. - The company framed the issue as a shift from identity-centric security to data-centric governance. - Fasoo AI said organizations need policies that travel with the data and define how sensitive information can be accessed, shared, transformed or used by AI regardless of where the data resides or which AI agent requests it. - The company said security teams should rely on persistent data protection based on content, sensitivity, business context and user privilege. - Fasoo AI said that approach can control how AI agents use enterprise information without slowing AI adoption or disrupting productivity. - Fasoo Data Radar, Wrapsody and Fasoo Enterprise DRM enforce policy tags and apply persistent protection directly to sensitive files and documents. - Fasoo AI said those controls stay with the data throughout its lifecycle, allowing organizations to define and enforce usage policies at all times. - Fasoo AI-R DLP adds AI-aware data protection intended to reduce the risk of sensitive information being submitted to external AI services or unauthorized AI applications.

Between the lines: - The company is positioning data governance as a required layer for enterprise AI, not an optional add-on. - The message reflects a broader security concern: if AI agents can act across systems, then controlling only login credentials leaves too much exposure inside the enterprise. - Fasoo AI is linking its product set to that shift by tying policy enforcement directly to the data rather than to the application alone.

What's next: - Fasoo AI expects data-centric governance to play a larger role as organizations expand AI agent adoption. - The company said extending security from systems to the data itself will help enterprises embrace Agentic AI while keeping control over valuable information assets. - More information is available in Fasoo AI's data-centric ACL management page. - Additional company information is available on Fasoo AI's website.