Cohesity has announced broader backup and recovery coverage across Microsoft environments, alongside a way to make curated enterprise data available in Azure Databricks without copying it out of Cohesity Data Cloud. The practical draw is a wider safety net for Microsoft workloads and a more direct route from governed business data to AI and analytics.
Watch Desk analysis
What happened
The changes span Azure, Microsoft 365, DevOps and identity systems. Cohesity says its Azure protection now covers Blob Storage, Data Lake Storage and Kubernetes Service, with application-consistent backups and granular Kubernetes recovery. For GitHub and Azure DevOps, preview capabilities cover repositories and complete Azure DevOps organisations, with recovery options and an isolated, immutable copy.
For Microsoft 365, the company says it has added threat scanning across Exchange, SharePoint and OneDrive, and protection for the Preservation Hold Library. Its Gaia Catalog integration with Azure Databricks lets organisations find, curate and publish data for AI and analytics while the underlying data remains in Cohesity Data Cloud. Cohesity says role-based and file-level access controls also carry through to natural-language queries in Microsoft 365 Copilot.
Our top picks
- Protect more Azure data
Backup coverage now includes Blob Storage, Data Lake Storage and Kubernetes Service, with granular Kubernetes recovery. - Recover DevOps assets
Preview protection covers GitHub repositories and Azure DevOps organisations, including boards, wikis and pipelines. - Scan and recover Microsoft 365 data
Threat scanning spans Exchange, SharePoint and OneDrive, with preservation-library protection and metadata-aware recovery. - Connect curated data to Databricks
Gaia Catalog can publish datasets to Azure Databricks without moving or duplicating the underlying data. - Keep Copilot answers permission-aware
Cohesity says its existing integration reflects role-based and file-level access controls.
Why it matters
The Databricks integration gives the announcement a direct AI connection: teams can make governed enterprise data available for AI and analytics without building a new extract, transform and load pipeline to move it. That could simplify access to data while keeping the source in place. The expanded recovery features, meanwhile, address the less glamorous question of whether organisations can restore the data and systems their AI workflows depend on.
Our read
There is useful substance here, particularly for organisations already invested in Cohesity and Microsoft services. The AI angle is a data-access integration, not evidence of a new model or a guarantee that every dataset is suitable for AI. For buyers, the sensible next step is to check which features are generally available, which remain in preview, and how the access controls work in their own setup. Resilience is easier to appreciate before the restore button becomes the main event.
What to watch
- When the GitHub and Azure DevOps capabilities leave preview.
- How the Databricks integration handles access permissions in customer deployments.
- Which Microsoft workloads and recovery options are available across each Cohesity plan.
Discussion spark: Should enterprise AI data stay in its existing security boundary, as Cohesity says its Databricks integration allows, or is that less important than making the data easy to use?
Sources and evidence
- Cohesity Expands Cyber Resilience and AI Data Capabilities for Microsoft Environments – Yahoo Finance (13 November 2026, 01:00 UTC)
- Trained to betray: How post-training supply chain attacks hijack coding agents – Microsoft (13 November 2026, 01:00 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.