Startups building at the intersection of AI and security have raised roughly $855 million across more than 150 seed rounds so far this year, putting the category on pace for a record, per the funding roundup. This week alone, Act Security emerged from stealth with $60 million disclosed across seed and Series A, and Hush Security added a $30 million Series A to govern non-human identities and enterprise AI agents.
The read is that AI creates the security category it then has to defend. Every enterprise deploying AI agents has multiplied the number of things that can act on its systems, and those agents are non-human identities: credentials and permissions operating faster than any person and often without direct oversight. Governing what an AI agent is allowed to touch, and proving it did only that, is a control problem that did not exist at this scale two years ago. Hush Security naming non-human-identity governance directly is the tell for where the money is going.
For operators, the 150-seed-round figure matters more than any single raise. A category with that many seed bets is one investors think is early and large, which means fragmentation before consolidation. Expect a wave of narrow tools, each securing one slice of AI-agent activity, followed by the platform vendors like Palo Alto and CrowdStrike acquiring the winners. The near-term problem is that the tooling to secure enterprise AI is being built at the same moment enterprises are deploying the AI, so the defenses lag the exposure. That gap is the risk, and it is why the capital is rushing in.
Bottom Line
AI adoption manufactured a new security category, and $855 million in seed money is chasing it. If you are deploying AI agents, assume the tools to govern them are immature, and treat non-human identity as the control problem it has become.