Veeam Software has announced the general availability of Veeam Intelligent ResOps for Microsoft 365, a tool designed to show enterprises what AI agents are doing inside their Microsoft 365 data, as regulators in Singapore sharpen their focus on agentic AI safety.

The launch comes in the same week that Singapore’s Infocomm Media Development Authority (IMDA) and the AI Verify Foundation introduced a new toolkit to help companies test AI agents for safety and reliability, reflecting growing regional attention on whether AI agents are operating safely, reliably and within their intended permissions.

Most organisations can’t see what their AI agents are doing

According to Veeam’s Data Trust & Resilience Report 2026, 88 per cent of organisations are already using or piloting AI agents, yet only 7 per cent qualify as truly AI-ready. More than 82 per cent of enterprises have unknown AI agents operating in their IT environments today, and 95 per cent say data challenges have already slowed their AI progress.

Veeam Intelligent ResOps for Microsoft 365, the first generally available offering in the Veeam Intelligent ResOps (iResOps) portfolio, is built as an add-on for Veeam Data Cloud for Microsoft 365. It connects in a few clicks with no data migration or re-architecture required, giving organisations visibility into known and unknown AI agents, AI-driven file changes, and sensitive or regulated data across SharePoint, OneDrive, Teams and Exchange.

Precision recovery when an agent gets it wrong

The tool is designed to let security and IT teams discover Copilot and other AI agents, see which sensitive data each agent can reach, and close oversharing gaps. When an agent modifies or deletes the wrong files, teams can see exactly what changed, file by file, and undo those changes with precision rather than rolling back an entire system. It also identifies redundant, obsolete and trivial data that can be quarantined, reducing both storage costs and the potential blast radius of an incident, and lets authorised teams query data, access, agent activity and backup protection in natural language through built-in AI agents.

“For twenty years, the resilience question was simple: Can I recover? In the agentic AI era, the question is whether you can determine precisely what was affected, who or what touched it, and how to return precisely to a trusted state,” said Anand Eswaran, CEO of Veeam. “A recovery point tells you what you can restore. A recovery map tells you what to restore first, what changed, and what not to bring back.”

Krista Case, Principal Analyst at theCUBE Research, said the research underlying the launch shows persistent gaps in how organisations monitor agent activity, reconstruct actions and validate recovery, adding that connecting data intelligence and governance with resilience operations is becoming essential for organisations to adopt AI while maintaining trust in their data.

What’s next

Veeam is also opening private previews of Intelligent ResOps for the Veeam Data Platform and for Okta, extending the same model of contextual intelligence to broader infrastructure and identity systems. Further private previews for GitHub, Azure DevOps and Google IAM are planned for the fourth quarter of 2026.

“Customers don’t want silos of tools with conflicting sources of truth. They want visibility, controls, and resilience in one place,” said Rehan Jalil, President of Products and Technology at Veeam.