Varonis Systems, Inc., a leader in data security, announced the addition of prompt monitoring to its Microsoft 365 Copilot security offering. This new feature allows security teams to investigate risky prompts, responses, and sensitive file references directly within the Varonis Data Security Platform.
Generative AI tools can dramatically improve productivity, but attackers and rogue insiders can also weaponize the technology to gather and steal sensitive data quickly. The ability to quickly identify and mitigate prompt hacking is essential for clearing security roadblocks currently holding up many AI projects.
Varonis for Microsoft 365 Copilot identifies suspicious prompts, such as employees searching for salary information or trade secrets. In addition to detecting blatantly malicious intent, Varonis uses machine learning to build behavioral profiles for each user and will detect when their actions — with Copilot or otherwise — become abnormal.
“Varonis has the unique ability to quickly detect and stop risky AI activity,” said Varonis Field CTO Brian Vecci. “Now, security analysts can replay the user’s entire Copilot interaction right within the Varonis UI to quickly get to the bottom of a Copilot-related alert.”
In addition to prompt monitoring and analysis, Varonis accelerates investigations with Athena AI — an embedded AI assistant that helps security analysts resolve incidents faster. Customers can also subscribe to Varonis’ Managed Data Detection and Response service to get 24x7x365 threat detection and response coverage with an industry-best service level agreement.
Varonis is a Top-Tier Microsoft Partner, providing security solutions that complement Microsoft’s rich ecosystem. Together, Varonis and Microsoft accelerate the safe adoption and use of Microsoft 365 Copilot for customers.
Organizations can request a free Varonis Data Risk Assessment to evaluate their AI readiness and start their Microsoft 365 Copilot journey directly from the Azure Marketplace.