Rubrik launches Agent Rewind to safely reverse AI agent errors
Rubrik has launched Agent Rewind, a tool designed to enable organisations to undo unwanted actions performed by AI agents.
The new capability, powered by Predibase's AI infrastructure, arrives after Rubrik's acquisition of Predibase was completed. Agent Rewind is intended to address errors made by autonomous AI agents, equipping organisations with greater transparency into agent activity and the means to revert changes to both applications and data.
AI agents, which act autonomously to perform or automate tasks, have demonstrated both the promise and the risk of such technology. Recent high-profile errors, including an incident in which an AI agent deleted sensitive data and attempted to conceal the deletion, have underscored concerns about agentic AI reliability. Studies continue to show that over half of AI projects fail to reach production, often due to unintended mistakes or results that do not meet enterprise needs.
Industry perspectives
Johnny Yu, Research Manager at IDC, commented on the need for solutions that address these risks. He said, "As companies consider investing in AI, they often don't take into account the mistakes that AI agents can and will make. Agentic AI introduces the concept of 'non-human error,' and as with its human counterpart, organisations should explore solutions that allow them to correct potentially catastrophic mistakes made by agentic AI."
Anneka Gupta, Chief Product Officer at Rubrik, explained how Agent Rewind is developed for the evolving AI environment. She said, "As AI agents gain autonomy and optimise for outcomes, unintended errors can lead to business downtime. Agent Rewind integrates Predibase's advanced AI infrastructure with Rubrik's recovery capabilities to enable enterprises to embrace agentic AI confidently. Today's organisations will now have a clear process to trace, audit, and safely rewind undesired AI actions."
These comments are mirrored by a recent study which found that agentic AI systems often become disoriented, take incorrect shortcuts, or fail to complete multi-step processes, raising reliability questions for their wider adoption within the enterprise.
Auditability and recovery
Agent Rewind is designed to provide an audit trail and create immutable snapshots, which allow organisations to reverse AI-driven changes if necessary. Traditional observability tools often show what happened but do not provide explanations or offer the ability to undo high-risk actions.
"Agent Rewind will close the loop on what happened, why it happened, and how to undo it," said Chad Pallett, 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."
The solution offers key features including context-enriched visibility, which records and contextualises AI agent behaviour and the impact of each action; safe rollback functionality utilising Rubrik Security Cloud for rewinding changes made to files, databases, configurations, or other repositories; and broad compatibility with multiple platforms and AI agent builders, such as Agentforce, Microsoft Copilot Studio, and Amazon Bedrock Agents, as well as with custom AI agents.
Ongoing AI challenges
The adoption of autonomous AI agents presents enterprises with new challenges in security, compliance, and operational integrity. The release of Agent Rewind follows increased scrutiny on AI agents after multiple incidents involving technical malfunctions, legal disputes, and disruption caused by erroneous or malicious actions performed by AI-driven systems. Visibility, traceability, and the ability to safely reverse changes are being increasingly recognised as necessary capabilities for organisations integrating AI agents into their workflows.
Rubrik positions Agent Rewind as enabling enterprises to achieve a higher level of trust and resilience as they adopt agentic AI, aiming to address both the technical and operational barriers that have historically prevented broader deployment and production use of autonomous AI within critical business functions.
