Jazz Secures $61M to Disrupt Data Loss Prevention with AI-Native Tech

The cybersecurity landscape is undergoing a significant transformation as organizations move away from rigid, rule-based security toward more fluid, context-aware systems. This shift was underscored this week by the announcement that Jazz, an AI-native security startup, has secured $61 million in funding to rethink Data Loss Prevention (DLP). For nearly two decades, DLP has been a polarizing category in enterprise security; while essential for compliance and intellectual property protection, legacy systems are frequently criticized for their high false-positive rates and the significant administrative burden they place on IT teams. Jazz enters the market with a promise to solve these systemic issues by applying large-scale machine learning and behavioral analysis to the problem of data exfiltration.

Traditional DLP solutions, such as those offered by incumbents like Broadcom (through its Symantec acquisition) and Forcepoint, rely heavily on regular expressions (regex) and manual data classification. This approach requires security teams to predict every possible format of sensitive data, from credit card numbers to proprietary source code. However, in the modern era of unstructured data and rapid collaboration, these static rules often fail. They either block legitimate business processes—causing significant security friction—or miss sophisticated exfiltration attempts that do not match a predefined pattern. Jazz's approach is fundamentally different, building an AI-native stack from the ground up rather than attempting to bolt AI features onto an aging architecture.

“

This shift was underscored this week by the announcement that Jazz, an AI-native security startup, has secured $61 million in funding to rethink Data Loss Prevention (DLP).

Jazz aims to solve the DLP dilemma by building a platform that understands the intent behind data movement. By leveraging large-scale machine learning models, the Jazz platform analyzes the context of a user's actions. For example, it can distinguish between a developer legitimately moving code to a secure repository and an employee attempting to exfiltrate intellectual property via a personal cloud account. This behavioral approach significantly reduces the alert fatigue that plagues Security Operations Centers (SOCs), where analysts are often overwhelmed by thousands of low-fidelity warnings generated by legacy DLP tools. The shift from reactive blocking to proactive, context-aware guidance represents a significant evolution in the security operations center toolkit.

A primary driver for this investment is the explosion of Shadow AI—the unauthorized use of generative AI tools like ChatGPT or Claude within the enterprise. As employees increasingly turn to these tools to summarize documents or debug code, they inadvertently risk leaking trade secrets or personally identifiable information (PII) into public AI models. Traditional web filters are often too blunt a tool for this problem, either blocking AI entirely or allowing it without oversight. Jazz’s AI-native architecture is designed to inspect the content and intent of these interactions in real-time, providing a more granular level of protection that allows for the safe adoption of AI productivity tools.