Upwind has raised $450 million in a Series C round and launched an AI lab focused on cybersecurity, as the company expands its security products for a world of AI agents. The funding is substantial; the more immediate question is what the company plans to build with it.
Watch Desk analysis
What happened
PYMNTS reports that Upwind will use the funding to support product and engineering growth. CEO Amiram Shachar said the company is building AI sensors for endpoints, agents and skills, and reworking its inventory graph.
Shachar also described three AI agents for customers: Blue investigates logs and other data to build an attack narrative; Red checks exposure and attack surfaces; and Green identifies areas where security teams can remediate findings. Upwind has also launched AI Labs for Cybersecurity, which Shachar says will work on security problems important to the industry’s future.
Why it matters
The announcement puts money behind a clear thesis: as AI spreads across company systems, security tools will need to help teams make sense of more endpoints, agents and findings. Upwind’s proposed agents cover investigation, exposure checks and remediation, rather than treating security as a single alerting problem.
That is a sizeable ambition, not proof the tools will improve security outcomes. The useful test will be whether they help teams prioritise and fix genuine risks without adding another layer of noise.
Our read
The $450 million round gives Upwind room to pursue that vision, while the three-agent setup gives it a concrete shape. The pitch is easier to assess than a generic promise to “use AI for security”, but deployments and results will matter more than the colour-coded agent names.
What to watch
- What Upwind’s AI Labs publishes or builds for cybersecurity.
- Whether customers can show that the agents improve investigation or remediation.
- How the company explains the agents’ access, oversight and role in security decisions.
Discussion spark: Should AI security agents be allowed to remediate findings automatically, or should every fix require a human sign-off?
Sources and evidence
- Source update (28 October 2026, 14:58 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.