Estimated reading time: 4 minutes · Last updated:
Upwind Security has closed a $400 million Series C round, led by Bessemer Venture Partners, to build out its cloud and AI security platform. The raise takes the company's total funding past $800 million and lands less than nine months after its $250 million Series B. Upwind says the money will go into product development and a larger engineering team, with new tooling aimed at endpoints, code and cloud environments. The Upwind Series C funding also backs three AI agents for security teams and a new research arm, AI Labs for Cybersecurity, that will explore four separate areas of study.
AI has a long reach within the enterprise environment, so we're building AI Sensors for endpoints, Agents and Skills scanners and are reinventing our inventory graph.
Amiram Shachar, CEO, Upwind
Key takeaways
- The round: Upwind Security raised $400 million in a Series C led by Bessemer Venture Partners, lifting its total funding above $800 million.
- Investors: Every existing Upwind backer, from Greylock Partners to Leaders Fund, returned for the Series C, and Canapi Ventures joined as a new investor.
- Pace: The Series C arrived less than nine months after Upwind's $250 million Series B.
- Product plan: CEO Amiram Shachar says Upwind is building AI Sensors for endpoints and scanners for Agents and Skills, and reworking its inventory graph.
- Research: Upwind launched AI Labs for Cybersecurity to study four areas, including agentic security and frontier models.
Table of contents
Who Backed Upwind's $400 Million Round
Bessemer Venture Partners led the Series C. Every existing Upwind investor returned: Cyberstarts, Greylock Partners, Salesforce Ventures, TCV, Craft, Alta Park, Swish Ventures, Picture Capital, Penny Jar, Cerca Partners and Leaders Fund. Canapi Ventures came in as a new investor.
Upwind announced its $250 million Series B less than nine months before this round, so it has raised roughly $650 million across its last two rounds.
What Upwind's Blue, Red and Green Agents Do
Upwind has built three AI agents to help security teams investigate threats, confirm vulnerabilities and act on findings, and the company says customers already run them.
The Blue Agent works on live attacks. It pulls logs and other data together into a single view of what happened, so responders can trace and answer exploitation attempts at runtime.
The Red Agent checks whether a vulnerability or exposure can be exploited in a customer's environment, and teams rank fixes on those results. The Green Agent hunts for changes that would clear large numbers of findings across affected resources at once.
Where the New Money Goes Beyond the Agents
Amiram Shachar, Upwind's CEO, framed the product plan around AI's spread through companies. "AI has a long reach within the enterprise environment, so we're building AI Sensors for endpoints, Agents and Skills scanners and are reinventing our inventory graph," he said in the company's announcement.
Upwind also wants to deepen its identity, data and AI security features. Its research effort, AI Labs for Cybersecurity, gives researchers infrastructure and latitude to study four areas: advanced AI systems, agentic security, AI security, and frontier models, systems and benchmarks. The lab targets both present problems and future security needs.
| Agent | What it does | How teams use it |
|---|---|---|
| Blue Agent | Combines logs and other data into one attack picture | Investigate and respond to runtime exploitation |
| Red Agent | Tests if vulnerabilities and exposures are exploitable | Prioritise fixes on test results |
| Green Agent | Finds changes that clear many findings at once | Address risk at scale |
What could help or hurt Upwind's plan
The case for
- Full participation from all existing investors signals continued insider support.
- Customers are already using the Blue, Red and Green agents, according to Upwind, so the product base is not starting from zero.
- More than $800 million raised in total funds a broad push into endpoints, code and cloud.
The case against
- Shachar's AI Sensors and Agents and Skills scanners were announced without release dates.
- Expanding into identity, data, endpoint and AI security at once spreads engineering effort across many fronts.
- Two large rounds within nine months raise expectations for rapid delivery.
What to be careful about
- The agents' effectiveness rests on Upwind's own description; no independent results are given.
- No timeline was given for AI Sensors, the scanners or the rebuilt inventory graph.
- AI Labs for Cybersecurity has no stated outputs or schedule against which to judge it.
The bottom line
Two large rounds inside nine months give Upwind, now funded beyond $800 million, room to chase a broad brief: endpoints, code, cloud, identity and data, all framed around AI. The open question is execution. Shachar has named products, including AI Sensors and scanners for Agents and Skills, but Upwind attached no release dates to them, and gave no timeline for AI Labs for Cybersecurity. The agents are the most concrete piece, since Upwind says customers use them now. What the company ships with this $400 million, and how quickly, will show whether its investors' unanimous return was a sound call.
What to watch
- Watch for Upwind to release its AI Sensors for endpoints and its Agents and Skills scanners; the company has not set a date.
- Watch for the first research output from AI Labs for Cybersecurity; no schedule has been announced.
Frequently asked questions
How much funding has Upwind Security raised in total?
Upwind has raised more than $800 million in total after its $400 million Series C. That includes a $250 million Series B announced less than nine months earlier.
What will Upwind spend its $400 million Series C on?
Upwind says the money will support product development and engineering growth. CEO Amiram Shachar named AI Sensors for endpoints, Agents and Skills scanners and a reworked inventory graph, alongside broader identity, data and AI security features.
What does Upwind's AI Labs for Cybersecurity research?
AI Labs for Cybersecurity covers four areas: AI security, agentic security, frontier models, systems and benchmarks, and advanced AI systems. Upwind says it addresses both current challenges and future security needs.
Related reading
