Upwind Raises $400M for Cloud AI Security

Cloud security teams often face thousands of alerts without knowing which risks are genuinely exploitable. Upwind has raised substantial new capital to build security around what applications actually do while they run.

What happened

Upwind announced a $400 million Series C led by Bessemer Venture Partners, with participation from Greylock, Cyberstarts, TCV, Salesforce Ventures and other backers. The round values the cloud security company at approximately $3.8 billion, compared with a reported $1.5 billion earlier in 2026. Upwind says its cumulative funding now exceeds $800 million.

Its platform uses runtime sensors to observe behaviour across cloud applications, infrastructure and workloads. The goal is to distinguish meaningful exposure from configuration issues that may have limited real-world impact. Upwind is also extending its capabilities to cover AI workloads and agent-related security risks.

Why it matters

Modern applications operate across containers, cloud services and interconnected systems. Static scans can identify weaknesses, but they do not always reveal which ones attackers could realistically exploit. Runtime context can help security teams prioritise the vulnerabilities and suspicious actions that matter most.

AI adds another layer of complexity as organisations connect models and agents to sensitive systems. Monitoring the behaviour of those workloads could become an important enterprise requirement.

The valuation reflects investor expectations, not proof that every new feature has achieved broad customer adoption. Maintaining efficient growth and demonstrating reliable protection will matter as the business scales.

The bigger picture

Cloud security is shifting from counting possible vulnerabilities toward understanding live risks. Upwind's financing signals demand for platforms that combine infrastructure visibility with security for the emerging generation of AI-powered applications.