Cyberhaven Flow: A Data Security Platform Built for Human and AI Workers Sharing the Same Files

US security company Cyberhaven has announced “Cyberhaven Flow,” a data‑centric security platform designed for workplaces where both humans and AI agents are constantly touching the same information.

Instead of traditional DLP tools that simply block certain actions, Flow focuses on data lineage — tracking how a file is created, copied, summarized, or shared over time and across apps. The platform records who accessed which document, through which application, and where it was sent, whether that’s an internal system or an external AI chatbot.

When Flow detects sensitive material being pasted into an AI tool or moved to an unsafe destination, it can automatically block the action, require justification, or alert security teams. Cyberhaven says adoption of AI and “agentic” apps on employee devices has roughly doubled in the past year, and argues that companies are shifting from trying to ban AI to trying to use it safely. For organizations rolling out AI copilots and autonomous agents at scale, a security layer that understands both human workflows and AI‑driven activity is quickly becoming a core requirement.