Former Google Security Executives Launch AegisAI with $13 Million Seed to Combat Email Threats
Key Points
- AegisAI founded by former Google Safe Browsing and reCAPTCHA executives.
- Secured $13 million seed funding led by Accel and Foundation Capital.
- Deploys a network of autonomous AI agents to analyze emails in real time.
- Agents specialize in threats such as malicious links, attachments, QR codes, and behavior patterns.
- Claims up to 90% reduction in false positives compared with traditional solutions.
- Installation takes under five minutes via API for Google Workspace or Microsoft 365.
- Early customers include Lokker and Mesh Connect, with pilots in the U.S. and Europe.
- Plans to expand its agent roster to 50‑100 as threats evolve.
AegisAI, a new email security startup founded by former Google Safe Browsing and reCAPTCHA leaders Cy Khormaee and Ryan Luo, has emerged from stealth with a $13 million seed round co‑led by Accel and Foundation Capital. The company builds a network of autonomous AI agents that analyze every component of an email in real time to detect phishing, malware, and business‑email‑compromise threats. Early pilots in the United States and Europe have already added paying customers, and the team plans to expand its technical and go‑to‑market capabilities.
Founding and Funding
AegisAI was founded by Cy Khormaee and Ryan Luo, both former executives on Google’s Safe Browsing and reCAPTCHA teams. The duo raised a $13 million seed round, with Accel and Foundation Capital serving as co‑lead investors. The funding will support the expansion of technical talent and the development of a go‑to‑market infrastructure.
Problem Addressed
The startup targets the growing prevalence of email‑based attacks, noting that more than 90% of successful cyberattacks begin with a phishing email. Recent studies have shown that phishing messages generated by large language models achieve click‑through rates far higher than human‑written emails, underscoring the need for advanced defenses.
AI‑Driven Agent Architecture
AegisAI’s core technology consists of an orchestrating AI agent that identifies potential threats and then calls on specialized “buddy” agents. Each buddy is a custom‑tuned language model focused on a specific threat type, such as malicious links, attachments, QR codes, or behavioral patterns. The agents reason together in real time, producing a verdict that can automatically quarantine suspicious messages.
To date‑up, the company has built over ten agents and anticipates expanding the roster to 50‑100 as adversaries evolve. The autonomous approach differs from traditional rule‑based email security platforms by continuously self‑tuning to new variants of attacks, which the founders claim can reduce false positives by up to 90%.
Deployment and Early Adoption
Installation is designed to take no more than five minutes via an API for Google Workspace or Microsoft 365 accounts. After deployment, AegisAI provides an initial report detailing detected threats, false positives, and false negatives, then operates in read‑only mode for a week before activating quarantine.
The startup is already piloting its solution with customers in the United States and Europe and has secured three paying customers, including data‑privacy compliance software Lokker and crypto‑payment platform Mesh Connect.
Outlook
With a small team of six members and offices in San Francisco and New York, AegisAI plans to leverage its seed funding to broaden its technical expertise and scale its market presence. The founders believe that as adversaries adapt, their expanding network of AI agents will keep pace, offering a proactive defense against increasingly sophisticated email threats.