Island Series F: A $6.4B Bet on the Browser as AI Agent Control Plane
Island has raised $400 million in a Series F round that values the Dallas-based security company at $6.4 billion, a figure the company says is more than double its 2024 mark. The Island Series F was led by Evolution Equity Partners, with Sequoia Capital, Coatue Management, JPMorgan Growth Equity Partners, Prysm Capital, Cyberstarts and Insight Partners taking part. Island sells an enterprise browser that embeds security, data governance and AI agent controls into the workspace where employees and automated agents operate.
The round was announced on September 24, 2026. Island says the capital will fund the scaling of AI agent applications, the modernization of enterprise work methods, international expansion and additional hiring. The company describes its product as an agentic control plane for enterprises, a framing that places the browser at the center of how automated work gets authorized.
Why the Browser Became the Argument
Security teams have spent years pushing controls to the network edge and to the endpoint. Both vantage points have limits when the actor is an AI agent. Edge inspection has little to work with inside encrypted sessions, and endpoint tooling sees a process running without understanding the business task behind it. An agent that logs into a supplier portal, pulls pricing, drafts a purchase order and files it looks, at the network layer, like ordinary browsing.
The browser sits elsewhere in the stack. It handles the credential, the destination, the session and the payload in one place, which allows policy to be applied before an action executes instead of being reconstructed afterwards. For agents that drive web interfaces instead of clean APIs, that is often the only point where intent and action are visible at the same time.
Three questions have to be settled whenever an agent acts: which agent is acting, what it is allowed to do, and what it actually did. Identity systems answer the first for the human who started the session. The remaining two are decided at the moment of action, which is where a browser-level policy engine can intervene.
Agent identity itself remains unsettled. In most setups an agent inherits the access rights of the employee it acts for, so a misconfigured or compromised agent can reach anything that person can. A browser-level control allows a narrower grant: a session scoped to named domains, a time window and a list of permitted actions, all revoked once the task ends. That granularity is the commercial claim behind the Island Series F.
That argument gained commercial weight as enterprises moved agents out of pilots. Island's stated use of proceeds points at deployments touching many systems at once, and those are the deployments where one audit trail matters more than per-application integration work.
What the Capital Is For
Building and maintaining a browser costs more than shipping a typical software product. Distribution to large fleets requires compatibility engineering, IT procurement work and support for the policy controls that security teams demand before rolling anything out. Island has said the funding will support research and development, hiring and international expansion.
The investor list carries its own signal. Evolution Equity Partners led the round, and the participants mix early backers such as Sequoia Capital and Cyberstarts with later-stage growth investors including Coatue Management, Insight Partners, JPMorgan Growth Equity Partners and Prysm Capital. Growth capital at this scale usually implies expectations about enterprise revenue, not technology milestones.
At $400 million against a $6.4 billion valuation, the round sells roughly 6% of the company, and the doubling since 2024 implies the earlier mark sat below $3.2 billion. The new price assumes the category keeps expanding instead of settling into a niche.
Island also has to fund the unglamorous parts of enterprise software: certification reviews, regional data handling and support coverage across time zones. Those costs scale with customer count, which makes a raise of this size at this stage a distribution problem as much as an engineering one.
The company's own expansion language points beyond its original product. Governing agent activity across an enterprise, and not only inside a browser, implies integrations with the systems those agents touch. Integrations are a headcount expense rather than a licensing one.
Island's Series F at a Glance
| Item | Detail |
|---|---|
| Round | Series F |
| Amount | $400 million |
| Valuation | $6.4 billion |
| Lead investor | Evolution Equity Partners |
| Other participants | Sequoia Capital, Coatue Management, JPMorgan Growth Equity Partners, Prysm Capital, Cyberstarts, Insight Partners |
| Headquarters | Dallas |
| Valuation change | More than doubled since 2024 |
| Stated use of funds | Scaling AI agent applications, modernizing enterprise work methods, research and development, hiring, global expansion |
The Trade-Offs Against Other Control Points
Identity platforms position themselves as the natural home for agent permissions, since they already own the authentication event. Endpoint vendors control the machine an agent runs on. Network security stacks control the path its traffic takes. Island's answer is that none of those layers observes the full sequence of a browser-driven task.
Adoption is the cost of that approach. Employees have to work inside a managed browser, and every agent touching a web interface has to pass through it. That is a lighter integration burden than wiring policy into dozens of internal applications, but it changes daily tooling, and it forces Island to compete with browsers that arrive preinstalled on corporate laptops.
Deployment timelines are the practical test. Rolling a managed browser across a large organization runs in phases, and each phase surfaces application compatibility problems that must be resolved before the next wave begins. That schedule, more than the feature list, tends to determine when agent governance reaches production.
A second question is whether agent traffic keeps flowing through web interfaces at all. If more agents call APIs and tool endpoints directly instead of clicking through applications, the browser's vantage point narrows. Island's language about expanding from the enterprise browser into a broader control layer suggests an attempt to follow agents wherever they run.
Budget dynamics add another constraint. Many enterprises are consolidating the number of security tools they operate, so a new purchase has to displace an existing line item or make several others cheaper to run. Island's positioning as an agentic control plane, and not a browser security product, is an argument for the second outcome.
Valuation adds pressure of its own. A $6.4 billion price assumes the category moves from specialist purchase to standing line item in mainstream security budgets. Investors who joined at this stage will look for enterprise penetration that can be measured.
How Buyers Should Read the Round
For security leaders evaluating agent governance, the round removes one procurement risk. A vendor with fresh capital and named growth investors is easier to defend in a review than a startup in the same category with a shorter runway. The counterweight is dependency. Moving a workforce onto a managed browser creates a commitment that is harder to unwind than swapping a network appliance, because it touches the interface employees use every day. Buyers should test the exit path as carefully as the agent controls themselves.
Competitors face a different calculation. Island now has the balance sheet to fund the long enterprise sales cycle that browser deployment demands, which raises the cost of matching it feature for feature. The likely response is more positioning around agent governance from identity, endpoint and network vendors, with each claiming its existing control point is sufficient.
Why This Matters
The Island Series F is a wager that agent governance will be bought as infrastructure, not added as a feature to tools enterprises already own. If that holds, the question of which browser a workforce uses stops being an IT preference and becomes a security architecture decision. The capital behind the idea means buyers will have a funded vendor to evaluate, and rivals will have a reason to answer with control-plane claims of their own.
Related Articles
- Five Months, $435M: AI Agent Security and Governance Funding Becomes a Category
- Snorkel AI Series E Raises $350M as Expert Data Gets Priced Like Infrastructure
- Crusoe $30 billion valuation puts contracted GPU revenue to the test
✔Human Verified
Researched and cross-referenced against primary sources by the Bytevyte editorial team. This article was generated with the assistance of artificial intelligence and reviewed by the Bytevyte editorial team.