Case study
How HiBob Gained Complete Visibility Into AI-Driven Data Exposure Across All Endpoints
HiBob is where thousands of companies run HR, which means their platform holds some of the most sensitive employee data in any organization: personal records, compensation figures, performance history.
When AI tools started spreading across every team, the security team faced a new set of questions. Who's pasting customer data into ChatGPT? Is source code with embedded secrets reaching third-party services? Are departing employees moving files to personal accounts on their way out?
HiBob wanted employees moving fast with AI without giving up visibility into where sensitive data was going. That visibility was hard to come by: most security tools can't see inside the browser or the desktop apps where AI now lives, which is exactly where the risk was sitting.
"We wanted our employees to use AI tools to increase organizational efficiency. In order to allow it we had to obtain visibility into what data was leaving or moving through these tools."
Tamir Ronen, Global CISO
In just a few weeks, HiBob deployed Bold across all endpoints, enabling on-device data classification and real-time intervention at the intersection of human, AI agent, and data risk. Because the AI runs on the device rather than in the cloud, Bold is fast enough to act before sensitive data leaves — and content is never shipped off for analysis. Bold's lightweight AI agent drove two use cases: data loss prevention across AI tools, and managing risk when employees leave. Different problems, same foundation: full context at the endpoint, the moment data moves.
With Bold, HiBob enabled fast AI adoption across the organization without losing control of how sensitive data moves. Employees can use tools like Claude and ChatGPT freely, but only through approved corporate accounts. Bold prevents sensitive data from reaching unmanaged or personal AI accounts, whether that's source code from an IDE, screenshots containing restricted information, or internal data shared with ungoverned AI agents.
When an employee moves data through a restricted tool or account, Bold intervenes in the moment — coaching, redirecting, or blocking based on the context of the action — rather than alerting after the data is already gone.Because Bold has full context (the type of data, the identity of the employee, and the action being taken), it minimizes false positives and helps employees make the right call without disrupting real work. And crucially, all of this happens without pulling in the security team.
Departing employees are one of the highest-risk moments for any organization, and one where Bold's on-device approach shows its value. Previously, investigating where files went meant jumping between siloed systems to piece together fragments. With Bold, the security team gets a complete picture of each incident through deep lineage, data-at-rest analysis, and a full understanding of the data itself.
When a departing employee moved internal documents to a personal account, Bold automatically traced the entire flow, from the originating application and local file path to the destination account. The security team could investigate and respond immediately, with no manual reconstruction and no blind spots where network and cloud-based tools lose visibility.
"We don’t just get an alert saying 'file was shared.' Now we can see where the file came from, how it moved through the machine, and where it ended up. That's the difference between an alert and an investigation."
Amit Hajaj, Security Engineer
Within weeks of deploying Bold across their endpoints, HiBob had full visibility into how sensitive data moved, who moved it, and where it went. That visibility paid off across both use cases:
HiBob chose Bold because they recognized where enterprise risk is shifting: back to the endpoint. The moment a person or an AI agent interacts with data on a device is the moment that needs visibility, context, and the ability to act, and that is what Bold was built to provide.