AegisAI just pulled in $36 million to take on one of the fastest-growing threats in enterprise security: AI-crafted spear phishing. According to TechCrunch AI, the Series A was led by Battery Ventures, with existing backers Accel and Foundation Capital joining in. That brings the young startup’s total funding to $49 million, less than a year after it launched.
What makes this one worth watching is who’s behind it. AegisAI was founded by former Google security executives Cy Khormaee and Ryan Luo, the people who worked on safe browsing tech and reCAPTCHA. These aren’t newcomers guessing at the problem. They spent a decade defending the world’s most-used email system, and they built AegisAI around a specific bet: the old way of stopping email attacks is broken.
What’s actually changing
For years, email security ran on rule-based systems. If a message matches a known bad pattern, block it. That worked when attacks were crude and repetitive. It doesn’t work anymore.
Attackers now use AI to research a target and write a message tailored to them. TechCrunch AI reports that AI can instantly pull together details about your co-workers, your active projects, even your recent travel, then craft an email that reads as completely authentic. Khormaee put a hard number on it: “AI-powered attacks bypass existing controls more than half the time now, which means they’re almost twice as effective as they used to be.”
AegisAI’s answer is to fight AI with AI. Instead of checking messages against a static list, its AI agents read each incoming email the way a careful human would, looking for small anomalies a checklist would never flag. Khormaee says the agents can catch threats legacy tools miss entirely, including malicious PDF attachments that look legitimate at first, some with built-in passwords and CAPTCHA designed to slip past standard spam filters.
Why it matters
The status quo here is a group of established vendors like Proofpoint and Mimecast, plus newer players like Abnormal Security. Their tools were built for a pre-generative-AI threat model. When the attacker has an AI writing bespoke messages at scale, a defense that reacts to known patterns is always a step behind.
That gap is exactly what investors are chasing. Battery Ventures general partner Dharmesh Thakker went looking for a startup that could defend against AI with AI after he noticed email attacks climbing. “The bad guys are using email to attack us using AI at a much faster pace than we can keep up with,” Thakker told TechCrunch AI. “Defending against that is going to be a number one priority for a lot of companies.”
Early demand backs that up. AegisAI says dozens of customers have already signed on, including crypto payments company Mesh, AI startup LangChain, and privacy compliance platform Lokker.
The competitive picture
AegisAI isn’t alone. Lightspeed-backed Ocean is chasing the same idea, using AI to analyze the context of every incoming email and spot impersonation attempts. So the “agentic email defense” category is starting to get crowded fast.
Thakker’s argument for AegisAI comes down to pedigree. A team that helped secure Gmail, he believes, has the best shot at becoming the leading new hack-prevention company. Whether that reputation converts into market share is the open question, but it’s a real edge in a space where trust is everything.
What to expect next
Email is the starting point, not the destination. Khormaee says AegisAI plans to expand into other areas of defense, such as data security, using the same core approach: custom AI agents that run investigations rather than match rules.
“The core idea of building customized, highly advanced agents that can do investigations is going to determine who becomes the next dominant security company,” Khormaee said.
For security teams, the practical takeaway is straightforward. If your email defense still runs mostly on if-then rules, assume attackers have already figured out how to walk around it. The market is moving toward AI that reasons about intent, and this raise is a clear signal of where the money expects the next few years to go. Full details are available at the original TechCrunch AI report.