With AI-generated email attacks up 5x in a single year and FBI-reported cybercrime losses hitting a record $20.8 billion, former Google reCAPTCHA and Safe Browsing leaders are scaling the first defense built for a world where no employee and no amount of training can spot the lure
SAN FRANCISCO, July 23, 2026 /PRNewswire/ — AegisAI, the email security company building its own large language models (LLMs) to defend the inbox, today announced a $36 million Series A led by Battery Ventures, with participation from existing investors Accel and Foundation Capital, to fight the next generation of increasingly sophisticated, AI-powered email attacks, which are having devastating consequences for individuals and organizations.
The round brings the company’s total funding to $49 million, less than a year after the company emerged from stealth. AegisAI will use the funding to scale its fleet of autonomous defense agents, accelerate general availability of Vanguard, its agent that hunts threats beyond the inbox, and expand enterprise go-to-market.
AI has enabled a new form of attack dubbed AI spear phishing. These attacks use LLMs to scan the internet for information about a victim and create highly targeted personalized lures designed to trick people into sharing sensitive data, downloading malware or sending money. This risk is turbocharged by the rising use of AI by individuals and workers. More widespread use of AI agents, which do work on someone’s behalf without human involvement, creates a huge new attack surface for criminals, who may be able to operate undetected for long periods of time.
AegisAI’s technology works by leveraging intelligent language models and an orchestrated network of AI agents to analyze and act on email threats in real time, essentially fighting AI attacks with AI. The company’s adaptive platform uses sophisticated reasoning to evaluate the specific intent of suspicious emails, instead of simply looking for patterns from past scams. This enables users to more quickly detect fraud, which is critical as the volume of email phishing has increased dramatically since the launch of ChatGPT.
“The most immediate, catastrophic risk to your organization isn’t an AI agent hacking your firewall. It’s an AI model manipulating someone in your organization into handing over the keys, often through the most trusted, most vulnerable contact of the person it’s targeting,” said Cy Khormaee, co-founder and CEO of AegisAI. “You cannot patch human trust. If your security program still relies on template-based phishing tests and awareness training, you are training your people to spot last year’s threat, not a capable agent crafting a novel lure just for them. When the attack is AI, the defense has to be AI.”
The future of phishing is perfection
For decades, the real defense against targeted spear phishing was the effort it required. Researching a target, mapping their relationships, impersonating a trusted vendor and timing the lure all took a skilled human operator – a capability reserved for nation-states. Today, that capability costs about the price of a cup of coffee. Off-the-shelf AI can autonomously pull public information on a human target, map their professional relationships, identify their most trusting contact, then craft a flawless, context-aware lure, at unlimited scale.
The evidence of spear phishing’s rise is now unambiguous. AegisAI’s State of the AI Threat in Email study, presented at the 2026 M3AAWG conference and based on analysis of more than 20,000 phishing, scam and malware emails, found:
- 5x growth in one year. AI-generated spear phishing grew from 2.8% to 13.9% of all observed phishing in 2025.
- 75% more effective. AI-generated emails evade traditional filters at nearly double the rate of human-written attacks, reaching the inbox more than half the time.
- Authentication is no defense. 72.6% of successful AI attacks passed email authentication, sent from compromised legitimate accounts with established sending histories.
The FBI’s 2025 Internet Crime Report tells the same story from the victim’s side. Reported cybercrime losses reached a record $20.8 billion last year. Phishing complaint volume stayed essentially flat – but the losses those complaints represent surged more than 200% in a single year, from roughly $70 million to more than $215 million. Precision has replaced volume. Business email compromise, payload-less attacks that exploit identity rather than software, accounted for $11.64 billion in losses, versus less than $52 million for ransomware and malware combined: a 365-to-1 ratio.
AI vs. AI
The security industry spends enormous energy debating advanced autonomous agents that chain minor technical vulnerabilities into major exploits. AegisAI argues the same capability has already arrived at the human layer and it is the more immediate threat.
AegisAI was founded in 2025 by Khormaee and Ryan Luo, veterans of Google’s core security group who helped build reCAPTCHA, Safe Browsing and Web Risk, systems that protect billions of users daily. Instead of scanning for known-bad signatures, AegisAI deploys autonomous AI agents inside the inbox that interrogate the intent and identity behind every message, catching linguistically perfect attacks that pass every technical check, while cutting false positives by up to 90% compared to traditional solutions.
With Vanguard, announced in March, those agents now extend beyond the inbox. When a message is flagged, Vanguard follows suspicious links and attachments across the open web just as a user would – defeating adversarial CAPTCHAs, cloaked pages and weaponized documents – and returns a complete threat report in minutes.
“Email is where enterprise trust lives and generative AI just broke every assumption legacy email security was built on,” said Dharmesh Thakker, general partner at Battery Ventures. “When attacks are machine-generated, personalized and indistinguishable from legitimate mail, the only viable defense is an equally capable AI operating at machine speed. Cy and Ryan spent a decade building exactly these systems at Google scale. They’re the team we’d back to win this market.”
Since its public launch in September 2025, AegisAI has deployed with dozens of customers across fintech and technology, including:
- Mesh, a crypto payments company, where AegisAI’s agents catch everything from fuzzing attempts to AI-generated spear phishing and business email compromise without the security team spending time managing rules.
- LangChain, one of the most visible companies in AI and a prime target for sophisticated, AI-generated attacks given its public-facing engineers and executives.
- Lokker, where AegisAI caught an attack that came through compromised Salesforce infrastructure. This threat had nothing to do with a bad link or attachment, just a trusted vendor’s own systems being used against them.
About AegisAI
AegisAI, the email security company building its own LLMs to defend the inbox, is headquartered in San Francisco. Founded by the team behind Google’s reCAPTCHA, Safe Browsing and Web Risk, AegisAI deploys autonomous AI agents that defend organizations against AI-crafted phishing, business email compromise (BEC) and evasive payloads that slip past traditional filters. With $49 million in capital raised to date, AegisAI is backed by Battery Ventures, Accel and Foundation Capital. Learn more at aegisai.ai.
SOURCE AegisAI