AvePoint launches AI data classification & recovery

AvePoint has launched Kinetic Classification and added new intelligence to Rapid Recovery, expanding its tools for data governance and restoration across cloud software used in AI deployments.

The new classification product is designed to continuously reassess data sensitivity as files, access rights and usage change, rather than relying on a one-off label. It works across Microsoft 365, Google Workspace and a wider set of business applications, including GitHub, ServiceNow, Jira, Confluence, Box, Okta, Smartsheet, Monday.com, DocuSign, Bitbucket and AWS S3.

Rapid Recovery has also gained features intended to help security teams decide what to restore first after a cyber incident. The additions include intelligent recommendations on priority data, a Rapid Recovery Wizard for building recovery plans in advance, and Express Recovery for Entra ID to bring identity systems into the recovery process.

The launches come as companies face growing pressure to control the data used by AI systems while also preparing for outages, breaches and accidental exposure. Channel partners are increasingly being asked to advise customers on securing data in cloud environments and keeping services running when systems are disrupted.

AvePoint tied the new products to research from its latest State of AI Report, which found a gap between confidence and outcomes. According to the company, 82.7% of organisations said they were "very" or "extremely" confident in their ability to prevent unauthorised data access, yet 72% of those in the "very confident" group still experienced an AI-related unauthorised access incident in the past 12 months.

Classification shift

AvePoint argues that static labelling no longer reflects how data changes in modern organisations, particularly when AI systems and automated agents can access information across several software platforms. In that environment, a document judged safe to expose at one point may become sensitive later because of changes in ownership, retention status, permissions or surrounding content.

Kinetic Classification is intended to address that by combining automated and AI-led classification across sensitivity and retention, then revisiting those judgements as new signals appear. That creates an up-to-date view of sensitive information rather than leaving teams to wait for manual audits.

John Peluso, Chief Technology Officer at AvePoint, said the launch extends the company's wider AI governance model. "These capabilities are great examples of how we enable AI trust within organizations all around the world. For 25 years, AvePoint has been the trusted layer beneath the world's most demanding data estates. Kinetic Classification and Rapid Recovery extend that foundation into the AI era, so organizations can classify, govern, and recover their AI estate with confidence. It's an important piece of what we call the Trust Layer for AI: a connected layer of governance, security, recovery, and backup controls that sits across an organization's data, creating the foundation for AI success," Peluso said.

AvePoint has framed the additions as part of a broader operating model that links classification with recovery planning. In practice, that means identifying which data is both sensitive and essential to business operations, then using that information to decide what should be restored first after an incident.

Recovery order

The issue has become more important as cyber attacks increasingly affect identity systems, collaboration platforms and file stores at the same time. Recovery efforts can stall if IT teams know which systems are down but not which data sets are most important to restore first.

AvePoint said its new recommendation engine is aimed at restoring what it described as the "minimum viable company", allowing organisations to bring back critical operations first. The Rapid Recovery Wizard is designed to let teams create and sequence recovery plans before an incident, reducing the need for manual triage during a crisis.

John Hodges, Chief Product Officer at AvePoint, said the challenge for many organisations is less about AI ambition than understanding the data underpinning it. "The next wave of AI governance starts with the data itself. Our 2026 State of AI Report makes clear that organizations aren't struggling with ambition; they're struggling with readiness: knowing what data is sensitive, who owns it, how it changes, and whether it should be used by AI in the first place. Kinetic Classification gives enterprises a living foundation for that work, continuously interpreting sensitivity, ownership, lifecycle, and risk so organizations can govern the data their AI is already using," Hodges said.

AvePoint said it serves more than 28,000 organisations and 6,000 channel partners across Microsoft, Google, Salesforce and other cloud environments.