Cohesity Bolsters AI Resilience with Data Governance, Federated Access

Event summary

  • Cohesity launched an 'Enterprise AI Resilience' strategy on March 10, 2026, focused on protecting AI infrastructure and enabling secure data access.
  • The strategy includes immutable snapshots and point-in-time recovery for AI environments, including agent memory and vector databases.
  • Cohesity introduced federated semantic search via the Model Context Protocol (MCP) to allow AI tools like Glean to access governed data.
  • The upcoming Cohesity Gaia Catalog will enable secure data access for analytics platforms like Databricks and Microsoft Fabric.
  • The strategy leverages Cohesity Data Security Posture Management (DSPM), powered by Cyera, for data governance and access control.

The big picture

As enterprises increasingly deploy AI models into production, the need for robust data security and resilience is becoming paramount. Cohesity's strategy addresses a growing market demand for solutions that protect AI infrastructure, govern data access, and enable secure data activation—a critical component for responsible AI adoption. The move to federated data access via MCP signals a shift towards more collaborative and secure AI ecosystems, potentially disrupting traditional ETL processes.

What we're watching

Agent Risk
The effectiveness of Cohesity's integrations with ServiceNow and Datadog in translating agent risk signals into automated recovery workflows will be a key indicator of the strategy's practical value.
MCP Adoption
The rate at which the Model Context Protocol (MCP) is adopted by other AI-powered tools will determine the extent of Cohesity's influence in shaping secure data access standards.
Gaia Catalog
The success of the Gaia Catalog in driving data access and adoption across analytics platforms will be crucial for demonstrating the value of Cohesity's unified data cloud.