As AI is increasingly helping hackers to start mass -scale email attacks, former Google Security leaders have joined forces to create autonomous AI agents aimed at stopping e -mail and business e -mail threats.
This is the mission back AegeanA new e -mail safety boot that just emerged from Stealth with $ 13 million in seed funding, led by ACCEL and the Capital Foundation.
More 90% of successful cyberattacks start With an e -mail “fishing”, according to the Federal Federal Security Service CISA CISA. A recent crowdstrike study (PDF) He also found that electronic fishing messages created by large language models (LLMS) had a click of 54% in 2024, much higher than 12% for human texts.
Aegisai aims to tackle this growing threat with the suite AI.
Founded by former browser and Recaptcha Cy Khormaee and Ryan Luo and Ryan Luo, the start offers an orchestrated AI dealers’ network in real -time inspection, analyzing and neutralizing the e -mail threats, without relying on a specific set of rules. This approach disputes standard e -mail platforms based on static rules and often require extensive user training.
“The sum of all bad is a PDF attachment to an email. This is always where all the attacks started and so I really wanted to solve this problem,” Khormaee said in an exclusive interview with TechCrunch.
Khormaee was head of the product and product management manager on Google for over five years until July 2023. During this time, he led the security team responsible for the protection of Google, its four billion users and four million websites, Recaptcha and the risk of tissue. It was also during this period he first met with Luo, who had spent almost a decade on Google and was part of the safe browsing team.
Google gave Khormaee firsthand the experience of building electronic fishing technologies, a deep understanding of the company’s perspective security and how to develop and escalate security businesses quickly, told Techcrunch.
Prior to Google, Khormaee founded the Intelligence Sales Contastic platform, acquired by SugarCRM in 2016. Later served as Vice President of Product Management in careful for over one and a half years until November 2024 before Aegisai began.
Aegisai has built reasoning agents, each of whom is a custom LLM tuned to a particular threat. As soon as the orchestration agent recognizes a threat or possible threat, he calls other agents on the network, which Khormaee is referred to as “friends”. These factors then run the analysis, reason between them, and respond to the orchestrated agent with verdict.
Agents perform real -time analysis of each message item, including links, attachments, metadata, QR codes and behavioral standards.
“What we know from building these tools on Google is what is all about an email you need to analyze? What are all the sources of data? Khormaee said.
While Aegisai has currently built more than 10 agents for this project, Khormaee told TechCrunch that there could be 50 to 100 agents over time, as opponents become smarter and are trying to fool the system.
“I fully believe that in two years, opponents will understand what we are doing. They will rebuild and attack what we do and then we will need to build more agents to stay in front of them,” he said.
Unlike a standard e -mail safety platform that uses a rules -based approach, these AI agents find a bunch of attacks and self -service for any possible variant of these attacks in real time, Khormaee said. The start has developed multiple AI models tailored to various threats and specific industries, including those business capital and financial services.
Along with the rapid detection of threats, Aegisai agents help reduce false positives by up to 90% compared to traditional solutions, starting claims.
It takes “no more than five minutes” for customers to install the Aegisai system in a Google Workspace workplace or Microsoft 365 via an API, per khormaee. Once installed, the start will send a reference to a few days with the details of what the system was found in the environment, including false positive and false negative. He will then run on reading only for a week and then activate the quarantine.
“It is so difficult without this technology to solve this very heterogeneous email problem,” Khormaee said.
The start, with offices in San Francisco and New York, is currently managing a pilot with customers in the US and Europe and has already added three paying customers, including compliance software with the privacy platform and the encryption platform. The start has a group of six members today.
With the new investment, Khormaee said the start plans to expand its technique of know-how and create a strong Go-To-Market infrastructure.
