As enterprises accelerate AI adoption, a growing challenge is emerging: understanding whether the data feeding AI systems remains appropriate to use as it changes over time.
To address this, AvePoint has announced Kinetic Classification, a new capability designed to continuously evaluate data sensitivity throughout its lifecycle, replacing traditional point-in-time classification with a dynamic approach that adapts as enterprise data evolves.
The launch forms part of AvePoint’s broader Trust Layer for AI, bringing together governance, security, backup and recovery capabilities to help organisations deploy AI securely across increasingly complex cloud environments.
Unlike conventional data classification, which typically assigns labels once and relies on manual reviews to remain accurate, Kinetic Classification continuously reassesses data as access permissions change, ownership shifts, content evolves and AI agents interact with information. The result is a more up-to-date understanding of what data is sensitive and whether it remains appropriate for AI systems to access.
The announcement comes as organisations continue to balance AI innovation with increasing governance and security demands.
According to AvePoint’s latest State of AI Report, 82.7% of organisations say they are “very” or “extremely” confident in preventing unauthorised AI data access. However, 72% of respondents in the “very confident” group still experienced an AI-related unauthorised access incident during the previous 12 months, highlighting a gap between confidence and operational readiness.
“The next wave of AI governance starts with the data itself,” said John Hodges, Chief Product Officer at AvePoint. “Our 2026 State of AI Report makes clear that organisations 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 combines AI-driven and automated classification across sensitivity and retention, continually refining classifications as new signals emerge, labels expire, policies change and AI systems engage with content. This allows organisations to identify newly sensitive information before it is inadvertently exposed to AI applications.
The capability also reflects a broader shift towards managing AI governance across multiple platforms rather than within individual productivity suites. AvePoint says Kinetic Classification extends visibility across Microsoft 365, Google Workspace and a growing range of enterprise applications including GitHub, Jira, Confluence, ServiceNow, AWS S3, Box, Okta, Smartsheet, Monday.com, DocuSign and Bitbucket.
Alongside the classification announcement, AvePoint introduced new intelligence within its Rapid Recovery platform to help organisations recover from cyber incidents more efficiently. New capabilities include intelligent recovery recommendations that prioritise business-critical data, a Rapid Recovery Wizard that enables pre-built recovery plans, and Express Recovery for Microsoft Entra ID, extending protection to identity services.
John Peluso, Chief Technology Officer at AvePoint, said these capabilities are designed to provide enterprises with a stronger operational foundation for AI adoption.
“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 organisations can classify, govern and recover their AI estate with confidence.”
As AI becomes increasingly embedded within enterprise workflows, governance is moving beyond static policies towards continuous oversight of data, identities and AI interactions. By pairing adaptive data classification with intelligent recovery capabilities, AvePoint aims to help organisations build the operational resilience required to scale AI while maintaining trust, compliance and security.
