AegisAI, a startup founded by former Google security executives Cy Khormaee and Ryan Luo, has raised a $36 million Series A round led by Battery Ventures, with existing backers Accel and Foundation Capital also participating. The round brings the company's total funding to $49 million, following a $13 million raise reported by TechCrunch in September 2025. (TechCrunch)

Khormaee and Luo previously worked on safe browsing technology and reCAPTCHA at Google. Their new company deploys AI agents, autonomous software programs that can actively investigate suspicious content, to detect and block AI-driven spear phishing attacks in email. Spear phishing is a targeted form of phishing aimed at specific individuals or organizations rather than a broad scatter-shot approach. Khormaee says these attacks bypass existing controls more than half the time. The core technical differentiator AegisAI claims is its agents' ability to intercept malicious payloads that conventional email security stacks miss, including PDF attachments that use built-in passwords and CAPTCHAs to evade standard spam filters. (TechCrunch)

The threat model is straightforward and getting worse. Generative AI has lowered the cost and raised the quality of targeted phishing to a degree that legacy rule-based and statistical filters were not designed to handle. When an attacker uses a large language model to craft a contextually rich, grammatically perfect message and wraps the payload in interactive elements that defeat static analysis, the defender needs a system capable of equivalent reasoning over the message content and its attachments. That is the gap AegisAI is building for.

Dharmesh Thakker, a general partner at Battery Ventures, led the investment. Named customers include crypto payments company Mash, AI startup LangChain, and Google-owned privacy compliance platform Lokker. (TechCrunch)

The competitive landscape is crowded. TechCrunch names Lightspeed-backed Ocean, Proofpoint, Mimecast, and Abnormal Security as competitors. Proofpoint and Mimecast are entrenched incumbents with massive installed bases in enterprise email security. Abnormal Security has built a strong position by applying machine learning to the email threat problem and has been well-funded in its own right. Ocean, backed by Lightspeed, is a newer entrant pursuing a similar AI-native thesis. AegisAI is betting that its founders' specific background in large-scale abuse detection at Google gives them an edge in agent-based defense. (TechCrunch)

The company plans to eventually expand beyond email security into data security, though the Series A proceeds appear focused on deepening the current product. The email-to-data-security expansion path mirrors a pattern seen in the broader security market, where vendors that establish a foothold in one control plane use it as a wedge into adjacent areas. Whether AegisAI can execute that transition is an open question, but the intent signals ambition beyond being a point solution for phishing. (TechCrunch)

The broader context here is Khormaee's claim that AI-powered attacks bypass existing controls more than half the time. If that figure holds up across broader deployments, it implies a significant gap in the current email security posture of most organizations. The existing vendor base is not standing still, and incumbents like Proofpoint and Abnormal are also incorporating AI into their detection pipelines. What AegisAI is betting on is that an agent-based architecture, where autonomous AI systems actively probe and analyze suspicious content rather than passing it through a classification model, will prove more effective against adversarial AI than the approaches currently deployed.

There is a conceptual symmetry worth noting: the same generation of AI that is making phishing more dangerous is also enabling more capable defense. The open question for AegisAI and its competitors is whether the attacker's advantage in generating novel attacks at scale can be offset by a defender's advantage in reasoning about those attacks in real time. The market is effectively placing bets on that question, and Battery Ventures' $36 million is one of them.

The customer list, while still early-stage, spans interesting verticals. LangChain sits squarely in the AI ecosystem, where employees may be more attuned to AI-related threats but also more likely to interact with novel AI-generated content in their workflows. Mash operates in crypto payments, a sector with a long history of being targeted by sophisticated social engineering. Lokker, being Google-owned, adds a degree of internal validation from the same company whose alumni founded AegisAI. (TechCrunch)

The funding environment for AI security startups has stayed active even as the broader venture market has tightened. The pattern of a $36 million Series A, led by a top-tier firm with strong existing investor participation, suggests that investors see a durable category forming around AI-native security tools rather than a temporary feature arms race. In this author's view, whether AegisAI specifically captures enough of that category to justify its valuation will depend on execution against well-funded incumbents and equally well-funded newcomers. The technology thesis is sound. The go-to-market challenge is the harder problem.