AI
AI
AI
AI sandbox escape exposes gaps in enterprise assistant security
An AI sandbox escape discovered in Microsoft 365 Copilot highlights the risks that emerge when AI assistants connect to sensitive enterprise data and cloud infrastructure. The finding shows how familiar attack techniques can take on broader consequences in AI-enabled environments.
Joe Hladik (pictured), head of Zero Labs at Rubrik Inc., built his team’s research practice around backup data. This year, that focus expanded to employees’ everyday use of Microsoft 365 Copilot, which has been widely adopted across large enterprises.
“No one’s looking at backup data,” he said. “We found it to be a viable place to find actual intelligence to act upon.”
Hladik spoke with theCUBE Research’s Krista Case at Black Hat USA, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed the AI sandbox escape Rubrik Zero Labs discovered in Microsoft 365 Copilot and what it means for defending AI agents. (* Disclosure below.)
Tracing an AI sandbox escape from Copilot to Azure
The exploit allowed the researchers to escape Copilot’s sandbox and reach backend infrastructure in Azure, potentially exposing files stored within a customer’s tenant, including SharePoint and OneDrive content, according to Hladik. Although Microsoft patched the specific vulnerability, the underlying technique could apply to other AI assistants.
“We were able to basically break out of the sandbox and gain access to the backend infrastructure in Azure,” Hladik said. “Now you can browse and have basically free access to [those files]. So it’s a major, major find.”
The finding also points to broader visibility and governance challenges surrounding AI agents. Rubrik Zero Labs research found that only 23% of security leaders have full visibility into the agents operating in their environments, while Rubrik unveiled new AI agent governance tools at Black Hat USA, according to Hladik.
“Agents are just bots with models,” Hladik said. “They’re a bot that asks a model, and then the model will tell them what to do, and then they act. It’s new, it’s cool, but at the same time, I’ve seen this before.”
Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of Black Hat USA:
(* Disclosure: Rubrik Inc. sponsored this segment of theCUBE. Neither Rubrik nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)
Photo: SiliconANGLE
A message from John Furrier, co-founder of SiliconANGLE:
Support our mission to keep content open and free by engaging with theCUBE community. Join theCUBE’s Alumni Trust Network, where technology leaders connect, share intelligence and create opportunities.
- 15M+ viewers of theCUBE videos, powering conversations across AI, cloud, cybersecurity and more
- 11.4k+ theCUBE alumni — Connect with more than 11,400 tech and business leaders shaping the future through a unique trusted-based network
Are you an AWS customer? Support SiliconANGLE financially by buying your AWS services from our Marketplace portal page and links: https://siliconangle.com/aws-marketplace/
About SiliconANGLE Media
SiliconANGLE Media is a recognized leader in digital media innovation, uniting breakthrough technology, strategic insights and real-time audience engagement. As the parent company of SiliconANGLE, theCUBE Network, theCUBE Research, CUBE365, theCUBE AI and theCUBE SuperStudios — with flagship locations in Silicon Valley and the New York Stock Exchange — SiliconANGLE Media operates at the intersection of media, technology and AI.
Founded by tech visionaries John Furrier and Dave Vellante, SiliconANGLE Media has built a dynamic ecosystem of industry-leading digital media brands that reach 15+ million elite tech professionals. Our new proprietary theCUBE AI Video Cloud is breaking ground in audience interaction, leveraging theCUBEai.com neural network to help technology companies make data-driven decisions and stay at the forefront of industry conversations.
