“Every old control is breaking, so everything’s up for grabs,” Mike Fey said in an interview. The chief executive of Island, a Dallas security company, was explaining why he believes the rise of AI agents, software that acts on a company’s behalf, has thrown the cybersecurity market open in a way he has not seen before.
Investors appear to agree. Island said Sept. 24 that it had raised $400 million in a Series F round that values the company at $6.4 billion, more than double its valuation from 2024. The round was led by Evolution Equity Partners, with participation from existing backers including Sequoia Capital, Coatue Management, Insight Partners and Cyberstarts, and a personal investment from the cybersecurity investor Dmitri Alperovitch. The raise brings Island’s total funding to well over $1 billion.
Fey’s resume runs through the biggest names in the industry. He was president and chief operating officer at Symantec and, before that, general manager and chief technology officer of McAfee. Amiga, his co-founder, is a veteran of Israel’s 8200 cyber-intelligence unit who led software security for Microsoft’s Technology Center in Tel Aviv. The pairing of an American operator and an Israeli technical founder is a familiar shape in cybersecurity, and Island’s backers cite it as part of the thesis.
The company began four years ago with a narrower idea: the enterprise browser. Fey, formerly president and chief operating officer of Symantec and before that an executive at McAfee, founded Island with Dan Amiga, an Israeli engineer who invented web-isolation technology and sold his previous startup, Fireglass, to Symantec in 2017. Island emerged from stealth in February 2022 with a Chromium-based browser that built security controls directly into the place employees work.
From there the product sprawled. Island added data-loss prevention, then a SASE network architecture that replaced traditional VPNs with zero-trust access, and now a control plane for AI agents, the layer where companies decide what an internal AI assistant is allowed to see and do. “Just like humans, agents need the right context, access, guardrails and accountability,” Fey said.
The technical pitch is consolidation. Island’s platform performs zero-trust network access by continuously checking a user’s identity and context before allowing a connection to a private system. Its data-loss-prevention rules sit where work happens, blocking clipboard transfers, downloads and screen captures that a policy forbids. The same controls, the company argues, apply to an AI agent as naturally as to a person, because both act through the same browser and the same systems.
The browser was Island’s founding insight: it had become the office, and the security stack built for a world of desktops and network perimeters no longer matched where work actually occurred. The company’s customers, banks, retailers and government agencies among them, use it to fence corporate data off from personal browsing and to enforce policy on web apps the company does not control.
The timing is not accidental. As enterprises deploy AI agents that can read files, send email and change records, security teams are confronting a problem their tools were never built for: a worker that is software, moves faster than any human, and can be misled by a prompt. Reuters reported this week that companies have begun budgeting separately for AI-agent security risk. Island is positioning itself as the company that governs that new workforce.
The threat Island sells against is concrete. An AI agent with broad permissions can be induced by a cleverly worded prompt to forward a document, reset a password or wire a payment, and it does so without the hesitation a human employee might show. Security teams have no standard way to see what agents are doing, and Island’s pitch is that its platform provides the visibility and the guardrails: who the agent is, what it may touch, and what it did.
The valuation trajectory tells its own story. Island raised $175 million at a $3 billion valuation in 2024, then $250 million at $4.8 billion in March 2025, before the latest round. Analysts said the doubling in less than two years reflects a market willing to pay a premium for any company that can plausibly own the emerging category of agent security, a category that barely existed when Island was founded.
The round also lands amid a wider fight over how fast AI should move. CNBC, in its interview with Fey, noted the debate over AI regulation in Washington and Silicon Valley. Fey framed the uncertainty as opportunity. “I’ve never seen anything like it,” he said. “It’s a very special time.”
Island’s bet is that the old divisions of the security industry, network, endpoint and identity, collapse when the user is an agent. The company sells one framework for both human and machine work, and its backers are wagering that the next dominant security company will be the one that owns the agent, not the desktop. Island says it is used by some of the world’s largest companies, though it discloses little about revenue. The raise, its sixth in four years, gives it the runway to keep hiring and to build out its agent controls before larger rivals, and a wave of startups chasing the same fear, can catch up.
