Cohesity Enhances AI Agent Recovery Capabilities for Managed Data
Cohesity has unveiled a new solution called Cohesity Agent Resilience, designed to identify, secure, and restore infrastructure supporting enterprise AI agents.
Introduction to Cohesity Agent Resilience
Cohesity Agent Resilience is part of the Cohesity Data Cloud platform, offering a centralized view of agents and their dependent states. The solution addresses gaps in recovery capabilities as AI agents increasingly handle critical tasks without human oversight.
Key Features of Cohesity Agent Resilience
The system safeguards agent infrastructure, including memory, configuration, credentials, and workflows. It uses snapshot architecture, immutable backups, and clean-room recovery to protect agent memory and configuration. Two core functions include point-in-time recovery for agents and protection of databases, file systems, and services managed by agents.
Integration and Availability
At launch, Agent Resilience integrates with Amazon Bedrock, with support for Microsoft and Google platforms planned. The solution is available to select customers, with general availability expected by year-end.
Autonomous Cyber Resilience Framework
Cohesity’s Autonomous Cyber Resilience model leverages agentic workflows to automate a five-step framework: data and identity protection, recoverability assurance, threat remediation, application recovery practice, and AI risk posture optimization.
Agentic Workflows and Recovery Processes
The framework shifts from manual plans to continuous discovery, protection, and adaptation. Teams can define objectives via Cohesity Copilot, enabling automated workflows to evaluate impact, identify threats, and prepare isolated recovery environments.
Vasu Murthy, Cohesity’s Chief Product Officer, emphasized that while detection identifies AI agent deviations, it cannot reverse changes. Cohesity Agent Resilience offers a targeted recovery path for agents and impacted resources.
Enhanced Protection for Sensitive Data
Cohesity has enhanced capabilities to automate protection for newly discovered sensitive data. Customers with Cohesity Data Cloud Enterprise Edition and Cohesity DSPM can trigger protection actions based on policy for unsecured data.
Cohesity AI Resilience Academy
Cohesity AI Resilience Academy introduces a learning path with a free course, “Foundations of AI Resilience with Cohesity,” distinguishing AI assistants from agents and highlighting risks as AI shifts from advisory to actionable roles.
Course Overview and Benefits
The free, self-paced course covers AI agent concepts, risks, and lifecycle gaps. Learners will understand how Cohesity Gaia Catalog and Agent Resilience address these challenges. Post-Catalyst attendees receive enrollment links, with completers earning a Credly badge.
Srikanth Kuntamukkala, Cognizant’s CIS Global Lead for Partner Sales, noted that as AI agents handle more critical tasks, recovery capabilities beyond monitoring are essential. The ability to restore agent states and data they modify could offer a practical risk management approach.
Conclusion
Cohesity’s solutions aim to close recovery gaps in AI infrastructure, ensuring organizations can detect, protect, and recover from unintended agent actions. With autonomous workflows and educational initiatives, Cohesity positions itself as a leader in cyber resilience for AI-driven enterprises.
