Endpoint Security Faces a New AI Reality
Alan Shimel speaks with Nati Hazut, founder and CEO of Bold Security, during Techstrong TV’s Black Hat 2026 coverage. The conversation focuses on endpoint security and why the endpoint is becoming more important again. Hazut argues that the industry has accepted weak endpoint controls for too long. AI is now forcing teams to rethink that approach.

Hazut explains that Bold Security grew out of lessons from his previous work in data security and cloud security. While working with customers, he saw a basic problem. A user could still move sensitive data from a production system, zip it and send it through a SaaS app. Legacy data loss prevention tools often failed to stop that risk. That gap helped shape Bold Security’s focus.
AI Brings Workflows Back to the Endpoint
The discussion highlights a major shift in enterprise computing. For years, security teams moved toward consolidation, cloud services and browser-based workflows. Now AI is changing that pattern. Users are running AI tools locally and handling sensitive data on their devices. That makes endpoint security a front-line concern again.

Hazut says this creates new risk. The problem is not only malicious employees. It can also be accidental exposure caused by new AI tools. If sensitive data reaches a local model or unapproved workflow, the organization may lose control without realizing it. Bold Security is designed to help teams see and manage that activity in real time.
Local AI Models Can Improve Control
Bold Security uses local AI models on the endpoint to classify data and analyze user actions. Hazut says this helps address two old problems in a new way. First, many tools still rely on pattern matching and regex. Second, traditional systems often struggle to understand user intent.

By running AI locally, Bold Security aims to reduce privacy, cost and third-party risk concerns. The platform can look across apps, data, users, actions and AI workflows. Hazut says the goal is not to add another narrow agent. The goal is to give teams one control point for endpoint security and data protection.
Endpoint Risk Needs a Broader Platform
Alan and Hazut also discuss how endpoint tools have evolved since the antivirus era. Older tools were often heavy, limited or focused on one type of threat. Today’s endpoint risk includes data movement, AI use, SaaS workflows and user behavior. That requires a broader model.

For security leaders, the takeaway is clear. Endpoint security is no longer just about malware or device protection. It is about understanding how people, data and AI interact on the device. Bold Security is betting that this shift will make the endpoint central to enterprise security again.