· Venture Capital Tracker · investment-strategies · 1 min read
AI Security Funding, May–June 2026: From Perimeter to Control Plane
AI-security financings clustered around data, identity, agent behavior, and production controls rather than another generic perimeter product.
AI-security funding is separating into four control problems: data access, identity, model behavior, and production monitoring. Treating the category as one leaderboard hides the buyer and the technical boundary each company owns.
How to read the category
| Control problem | Example | What to validate |
|---|---|---|
| Data access | Cyera | Data discovery, governance, and deployment scope |
| Agent identity | Obsidian Security | Non-human identity coverage and workflow integration |
| Defensive model layer | Corma | Buyer, signal quality, and security outcomes |
| Continuous testing | Horizon3 | Frequency of testing and remediation loop |
The useful conclusion
The strongest funding stories connect a new AI capability to an old security budget and a measurable failure mode. “AI security” is not enough; a founder needs to show which asset is protected, who pays, and what changes after deployment.
This page is a framework around selected financings, not a complete market map. Amounts remain as reported and are not added together.
Sources and next step
Use the linked deal pages for primary-source detail, then compare relevant investors in the directory.
- 2026-vc-news
- startup-funding
- ai-security
- +1 more
Editorial note: AI tools assisted with research, structure, or drafting. Venture Capital Tracker retains human editorial responsibility for factual accuracy, relevance, and source quality before publication.
