Intent-aware endpoint security that protects what a human or AI agent is trying to do

Ent was founded by Elias Manousos and Brandon Dixon, who previously built RiskIQ, acquired by Microsoft in 2021, and then worked on Microsoft Security Copilot. Their argument for a new endpoint company is that the last decade of detection assumed a malicious binary and a human operator, while the risky activity on a modern endpoint increasingly looks like ordinary work: a contractor exporting a customer list on their last week, an employee pasting source code into a consumer chatbot, or an AI coding agent acting on an instruction nobody reviewed. None of those produce malware, so Ent positions the product as intent-aware, reasoning about what a sequence of actions is for rather than what data it touched.

That framing puts prevention back in scope. Instead of alerting after an exfiltration, the platform is designed to recognise the intent forming and intervene at the point of action, covering both human users and the AI agents now operating with delegated access on the same machines. Customers apply it to insider risk, AI usage governance, data loss prevention, last-mile threats, and incident investigation, and the company says it is deployed at Global 2000 organisations in hospitality, financial services, and defence. Ent emerged from stealth in June 2026 with a $100M seed round, among the largest on record, led by Decibel with participation from Sequoia, Crosspoint Capital, Craft Ventures, Shield Capital, Felicis, In-Q-Tel, S32, and SineWave. The company is based in San Francisco.