Artificial intelligence is developing rapidly, and as businesses adopt generative AI security measures, they face new challenges in protecting sensitive data from unintended exposure and misuse.

Organizations struggle to control sensitive data as AI-powered platforms such as Microsoft Corp.’s Copilot and Salesforce Inc.’s Agentforce offer unprecedented access to information. The challenge lies not only in managing permissions, but in understanding how data is used, who has access and what unintended risks may arise, Brian Vecci (pictured), field chief technology officer of Varonis Systems Inc., noted.

Varonis’ Brian Vecci talks with theCUBE about generative AI for the Tech Innovation CUBEd Awards 2025 interview series.

“We’ve always taken a data-centric approach to security, and especially these days with generative AI tools,” he said. “We continue to innovate to help ensure that our customers can get the benefits of these new tools and technologies without introducing any new risks.”

Vecci spoke with theCUBE’s Christophe Bertrand for the Tech Innovation CUBEd Awards 2025 interview series, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed the security and compliance risks of generative AI in enterprise environments and how Varonis helps mitigate data exposure.

How generative AI security protects sensitive enterprise data

One of the biggest risks associated with generative AI is its ability to surface hidden or forgotten data within an enterprise. Vecci shared a striking example of a major Wall Street bank piloting Microsoft Copilot.

“They gave people on the trading floor access to Copilot for 365, and one of the first prompts that one of these users put into Copilot was a relatively benign question: ‘What stocks do our employees invest in?’” Vecci said. “What they expected was they would get a summary. Instead, what this user got was a huge table of thousands of rows of employee names, their social security numbers, account numbers and positions within their 401(k) accounts. It’s because somebody on their compensation team had created a spreadsheet, saved it in a team site and had shared it inadvertently with everybody in the company.”

Varonis leverages AI-driven monitoring and automation to help enterprises securely deploy generative AI tools while minimizing data exposure risks, according to Vecci. In recognition of its contributions to enterprise security, Varonis received a CUBEd award for “Top Data Protection Innovation.” By leveraging generative AI security tools such as Varonis, enterprises can proactively monitor, detect and respond to data threats in real time.

“What makes Varonis unique and why we’ve been able to innovate in this way is that we don’t just know what the data is,” Vecci said. “We know what data is being used by who. When we build automation, and our customers apply that automation, they do it safely. We minimize how long it takes to detect a threat and how long it takes to respond to it. With our SaaS offering now, which is part of how our organizations are deploying Varonis, we see the issues before you do. So with managed data detection response, we’ll call you.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE Research’s coverage for the Tech Innovation CUBEd Awards 2025 interview series:

Photo: SiliconANGLE