Cohesity says it has been granted a US patent covering technology used in its generative AI platform for enterprise data, Cohesity Gaia. The company said the US Patent and Trademark Office issued Patent No. 12,619,501 on May 5, 2026, titled “Data Retrieval Using Embeddings for Data in Backup Systems.”
According to the company, the patent covers a method that combines secondary data systems—such as backups—with a retrieval-augmented generation (RAG) semantic layer intended to support GenAI applications. Cohesity said the approach is designed to allow GenAI to work with secondary data without replicating sensitive information into separate AI environments, while retaining existing governance, compliance, security and access controls.
The company said the patent was invented by Gregory Statton, Sanjay Poonen, Mohit Aron and Apurv Gupta. It described the core innovation as making secondary data “securely searchable” and usable as a governed source for GenAI applications, with AI workloads able to run against protected secondary data without moving or duplicating it.
In a statement, Cohesity CEO and president Sanjay Poonen said: “Protected recovery data is a goldmine. It is an organisation’s most important, complete, and trusted repository of enterprise information and institutional knowledge. Yet it remains among the most underutilised.” He added that the architecture applies AI directly to that data “without forcing organisations to move or duplicate sensitive information.”
Gregory Statton, listed as an inventor and Cohesity CTO and VP for APJ, said: “By enabling AI-driven insights directly from backup data, without moving or duplicating it, we’ve built something genuinely unique: a patented RAG architecture that keeps data in place.”
Cohesity also included a customer reference from the Netherlands. Patrick Ringelberg, Domain Data Centre Architect & AI at the Ministry of Infrastructure and Water management (Rijkswaterstaat), said: “In evaluating enterprise AI approaches, preserving our security posture and ensuring sovereign, on-premise control as a Dutch government institution were critical objectives,” adding that Cohesity’s approach “was the only one that made AI viable using our existing backup data as the foundation.”
The announcement comes as organisations look for ways to apply large language models to internal information while limiting data movement and potential exposure. By focusing on secondary data—often retained for recovery and compliance—vendors are positioning backup environments as inputs to AI-driven search and analysis, raising questions for security teams around governance, access control and the expanding set of systems that may be queried by AI.
Cohesity said Gaia is available as part of its Cohesity Data Cloud platform.
