Rubrik Debuts Agent Rewind, Addresses Agentic AI Errors

Data resilience and security leader Rubrik launched a new solution this week aimed at reducing the impact of agentic AI gone awry. We spoke with Rubrik’s Chief Product Officer, Anneka Gupta, about what she thinks the solution solves for organizations worldwide.

Observability in the AI era: Solution to when AI agents make mistakes

The announcement comes months after Rubrik closed its acquisition of Predibase, and the Agent Rewind solution is built on Predibase’s AI infrastructure. Rubrik says the product offering provides the following for users:

• Context-Enriched Visibility: Surfaces agent behavior, tool use, and impact while contextualizing each action, mapping it back to its root cause – from prompts to plans to tools – to enable precise recovery when something goes wrong. 
• Safe Rollback: Uses Rubrik Security Cloud to rewind what changed, whether that’s files, databases, configurations, or repositories. 
• Broad Compatibility: Will integrate seamlessly with a wide range of platforms, APIs, and agent builders, including Agentforce, Microsoft Copilot Studio, and Amazon Bedrock Agents, and will be compatible with any custom AI agent. 

“The genesis of this came as we were exploring how to leverage agentic AI internally,” Gupta said. “We began to see a problem in that AI agents can show such remarkable power and are so outcome-driven that it can be hard to put the proper guardrails in place to both leverage the productivity gains and mitigate the risks as much as possible.”

“Agent Rewind will close the loop on what happened, why it happened, and how to undo it,” said Chad Pallett, the chief information security officer at BioIVT, a global research partner and biospecimen solutions provider for drug and diagnostic development. “When using AI, there is a need for observability and secure rollback. Rubrik and Predibase will provide not just data safety and model speed, but also AI recoverability. In a market craving true observability and remediation, Agent Rewind is the answer I’ve been waiting for.”

Why Rubrik developed the ‘recoverability’ tool for the future

Rubrik says Agent Rewind makes “previously opaque AI actions visible, auditable, and reversible” by creating an audit trail and immutable snapshots that facilitate safe rollback in the event of mistakes. Most observability tools on the market currently focus on identifying the issue, but not on how to undo the incorrect actions. Rubrik also touts the integration capabilities built into Agent Rewind and its entire portfolio of solutions.

“We’ve designed this solution to integrate with existing frameworks because we know our customers already have their tech stacks built and they are already in these existing ecosystems,” Gupta said. “Our approach to this has always been to say, let’s go integrate with the players who are the best at what they do. We’re always focused on our customers having the flexibility to use our products within their existing frameworks. In that way, we’re future-proofing the solution for what might come next.”

Gupta acknowledges that Rubrik’s customer base, which is comprised almost entirely of large enterprises, represents a wide variety of AI adoption. While some are already considering agentic AI prototypes, others are nowhere near agentic deployment. For Gupta, that means Rubrik’s right where it should be in terms of its solution roadmap.

“We want to be ahead of where our partners are so that we can be a true thought partner with them on this AI journey,” Gupta said. “Agent Rewind will help our partners and customers who are ready for agentic AI now, and it will also encourage those who aren’t yet there to think ahead about the risks they will need to mitigate.”

What partners and end-users need to keep in mind when deploying agentic AI

Rubrik’s position on agentic AI is clear in how it’s marketing this offering: non-human mistakes can be just as costly as those of the (more human) past, and organizations need to have a response plan in place for when things go wrong.

“For us, it’s always about our heritage as a company, and that heritage is recovery. This is the next evolution in how we help customers recover, and what they are recovering from,” Gupta said.

“As companies consider investing in AI, they often don’t take into account the mistakes that AI agents can and will make,” said Johnny Yu, research manager at IDC. “Agentic AI introduces the concept of ‘non-human error,’ and as with its human counterpart, organizations should explore solutions that allow them to correct potentially catastrophic mistakes made by agentic AI.” 

Gupta also says she and other leaders at Rubrik are constantly talking with customers and soliciting their feedback. She highlights the speed at which AI agents will likely overwhelm internal security teams and other professionals– and hopes that Agent Rewind will be a key part of keeping enterprises secure as they embrace innovation.

“The amount of activity happening because of agents is going to massively dwarf the amount of activity by humans just because of the sheer level of productivity these agents can achieve, and I don’t think that’s years into the future, I think it’s maybe 12 months or so from now that we’ll start to see that,” Gupta said. “People are going to be overwhelmed, and we’re already hearing from CISOs and CIOs that they have a hard time getting their organizations to build the security foundation they need. Every CIO and CISO needs to be thinking ahead to this world.”