News
Commvault Pushes AI-Powered Recovery as Cyber Risks Accelerate
- By Gladys Rama
- November 17, 2026
Artificial intelligence is creating a new challenge for enterprise disaster recovery: the same technology that promises to improve IT operations is also making cyberattacks more effective and introducing risks from autonomous agents that can disrupt critical systems. In a press briefing this week ahead of its annual Shift conference, Commvault CEO Sanjay Mirchandani said, "AI-enabled recoveries are the way of the future." He argues that meeting those threats will require organizations to move beyond traditional backup and recovery practices toward faster, more comprehensive approaches to business resilience.
At its annual Shift conference, the company outlined how it plans to support that transition through new products, expanded integrations and AI-driven recovery capabilities. Announcements included Recovery Studio, designed to restore applications and their dependencies at machine speed; enhancements to AgentRecover, including expanded auditing and support for Microsoft Foundry and AWS Bedrock; and deeper partnerships with Nvidia, Zscaler and Databricks.
Together, the initiatives reflect Commvault's contention that AI-enabled recovery will become increasingly important as enterprises contend with both malicious attacks and unintended damage caused by AI systems.
All these changes are meant to use AI to address challenges to business continuity caused by AI, Mirchandani said during the keynote.
"When you invent the ship, you also invent the shipwreck," he said. "AI is no exception."
There has been no shortage of headlines about AI shipwrecks in recent months. In October, researchers used an OpenAI model to discover two critical security flaws in PHP, the language underpinning millions of webites. The month before, Microsoft identified an AI-powered threat actor dubbed Storm-3168 identifying Azure environments with vulnerable service principals to delete Azure Storage accounts rapidly and at scale. And this past summer, an agent running on Claude's Opus model independently acted to delete the entire production database of PocketOS, a provider of automation software for cars.
Together, they're emblematic of the multiple layers of AI-driven problems that enterprises are facing today, Mirchandani said. Malicious actors are using AI to make their attacks more effective; meanwhile, agents deployed internally are a risk all on their own when, able to perform destructive tasks because of a misconfiguration error.
"AI makes attacks faster [and] it makes mistakes faster," Mirchandani said.
Commvault is touting a "continuous recovery" model, powered by AI, as the new standard in security and resilience. As executives described it, this entails recovery across the full stack, going beyond data to include configurations, dependencies, policies and identities; an understanding of data's business context; and the assurance that the data being restored is "clean."
As Commvault CTO and Chief AI Officer Pranay Ahlawat said later in the keynote, "The smartest place for your dollar might be AI recovery."
Notably, Ahlawat cautioned enterprises against thinking recovery is less important than security.
"Security and recovery can't live in sepeate worlds anymore," he said.
Ultimately, executives said, organizations should aim to close the security-recovery loop -- and, increasingly, at machine speed.