Cybersecurity startup AegisAI has raised $36 million in Series A funding as businesses search for new ways to defend themselves against phishing attacks increasingly powered by artificial intelligence.
The company was founded by former Google security executives Cy Khormaee and Ryan Luo and is developing AI-based technology designed to identify highly targeted phishing attempts that traditional security systems may struggle to detect.
The funding round was led by Battery Ventures, with participation from existing investors Accel and Foundation Capital. The new investment brings AegisAI’s total funding to $49 million.
Phishing has long been one of the most effective forms of cybercrime.
Attackers send messages that appear to come from a trusted person or organisation and attempt to convince the recipient to reveal information, download malicious software or take another harmful action.
Spear phishing takes that idea further.
Instead of sending the same generic message to thousands of people, an attacker creates a message specifically designed for a particular person or organisation.
Artificial intelligence is making that process easier.
AI systems can rapidly analyse publicly available information about employees, companies, projects and professional relationships.
That information can then be used to create messages that appear far more personal and convincing.
An employee may receive an email that appears to reference a real colleague, an ongoing project or a recent business event.
The level of personalisation can make traditional warning signs less obvious.
AegisAI believes existing rule-based security systems are increasingly struggling to keep up.
Many traditional email filters depend on known patterns.
They may search for suspicious links, unusual language or previously identified malicious infrastructure.
AI-generated phishing messages can be created and modified quickly, making it harder for static rules to identify every new variation.
AegisAI is taking a different approach.
The company’s technology uses AI agents to analyse messages more contextually, looking for subtle signs that an email does not match legitimate communication.
The idea is to allow defensive AI to examine suspicious messages in a way that resembles human reasoning while operating at a much larger scale.
The company says its technology has already been adopted by dozens of customers.
That early interest reflects the growing anxiety surrounding AI-powered cyberattacks.
For businesses, phishing is no longer just a problem of employees receiving badly written emails claiming they have won a prize.
Attackers can now create highly convincing messages that mimic writing styles and reference real events.
The potential consequences can be serious.
A single successful phishing attack can provide access to corporate accounts, financial information or sensitive customer data.
From there, attackers may move deeper into an organisation.
The rise of AI has created something of an arms race.
Cybercriminals can use AI to create more convincing attacks.
Cybersecurity companies are responding by using AI to analyse and stop them.
AegisAI’s $36 million funding round is therefore part of a broader investment trend.
Investors increasingly believe that the next generation of cybersecurity companies will need artificial intelligence at the centre of their products.
The company’s founders also bring significant experience to the challenge.
Both Khormaee and Luo previously worked on security technology at Google, including systems connected to Safe Browsing and reCAPTCHA.
Those products have helped protect users from malicious websites and automated abuse for years.
Their new company is applying similar security experience to one of the most important threats emerging in the AI era.
The question now is whether defensive AI can evolve as quickly as the attackers using the technology.
That competition may define the next phase of cybersecurity.
For employees, the lesson remains the same.
Technology can provide protection, but people still need to be cautious.
An email that looks personal is not necessarily genuine.
And as AI makes cyberattacks more convincing, learning to question what appears familiar may become one of the most important security skills.
