Non-Human Identity and Transient Networks Extend Controls Beyond Human Workers
On September 24, 2026, a significant development in the realm of cybersecurity was reported. A startup, spearheaded by a former president of Symantec, secured $400 million in funding to enhance its pioneering worker-centric platform aimed at improving security for agentic workers. This endeavor, backed by Evolution Equity Partners in a Series F funding round, promises to extend control mechanisms across various facets, including data, identity, and observability—elements critical to modern cybersecurity.
The startup, named Island and headquartered in Dallas, has ambitions of not only achieving growth but also attaining profitability while preserving the option for an initial public offering (IPO). The company’s co-founder and CEO, Mike Fey, articulated the current landscape, asserting that legacy network designs, which have remained largely unchanged for a decade and a half, are inadequate for meeting the demands of the artificial intelligence (AI) era. This scenario presents a considerable opportunity for platform providers like Island, who are innovating to meet evolving security needs.
Fey conveyed his optimism regarding the timing of this fundraising initiative, expressing that it would empower Island to control its trajectory. "We can drive aggressive growth with this," he noted, "and propel ourselves into profitability, right through the IPO window. It grants us true optionality regarding our future direction."
Since its emergence from stealth mode in February 2022, Island has amassed a formidable workforce of 1,327 individuals and a hefty total of $1.05 billion in funding, with a substantial growth trajectory highlighted by a recent $250 million Series E round led by Coatue Management. Following the Series F funding, Island’s valuation soared to $6.4 billion, marking a 33% increase from the $4.8 billion valuation assigned during its previous fundraising round just a year and a half earlier.
An essential aspect of Island’s evolution has been its transition from serving traditional enterprise browsers to incorporating broader functionalities aimed at securing agentic workers—those leveraged by AI to function like conventional knowledge workers. By introducing control planes that encompass data, identity, observability, and network security, Fey envisions a vast total addressable market, poised to be supported by a comprehensive product line tailored for this new class of worker.
Employing sophisticated terminology, Fey characterized the workspace of the future, stating, "Once you have that, you now have that agentic agent surrounded by everything it needs to do its job well." In this envisioned future, Island’s infrastructure strives to provide a robust framework for AI agents—termed agentic workers—that necessitates comprehensive controls encompassing least-privilege access, effective data management, and identity solutions tailored for non-human actors.
Moreover, Fey elaborated on the importance of observability, which includes in-depth visibility into the actions of agents, their performance metrics, expenditure, and operational call graphs. Impressively, he mentioned that up to 90% of Island’s AI solutions are built upon their prior expertise in facilitating knowledge work.
Highlighting the operational challenges faced by agentic workers, Fey referred to their distinct needs—requiring not just the same tools as traditional knowledge workers, but also advanced capabilities like large language model (LLM) touchpoints, headless browser automation, transient least-privilege networks, and the management of non-human identities. Given that agentic workers often operate on servers and require adaptable deployment models and policies, he emphasized that Island’s architecture maintains scalability, mitigating the need for extensive rewrites or legacy system hindrances.
Fey pointedly remarked that, at this juncture, cybersecurity isn’t their primary concern. Instead, the focus lies in enabling capabilities that empower users to harness innovative technologies effectively and realize returns on investment critical for fostering the subsequent stages of AI adoption.
Transitioning towards the integration of AI technologies, Island employees are now witnessing a remarkable shift in operational methodologies. As clients increasingly express a desire to deploy advanced AI models like Claude safely, ensuring efficient model transitions and contextual protection becomes imperative. In light of this, Fey emphasized the necessity of maintaining a secure environment for sharing internal artifacts and establishing robust coding frameworks.
As the company’s journey continues, it became evident that Island is witnessing a dramatic shift from entirely manual coding practices to an impressive proportion of AI-generated code, which has surged from 0% to nearly 95%. This rapid transition necessitated the development of evaluators and judges to uphold quality assurance and leverage advancements from partners like Anthropic. However, Fey acknowledged the pressing need for additional engineers to create enterprise-grade modules across Island’s various control planes.
In the face of competitive pressures, particularly in network-centric deals where contenders like Zscaler and Palo Alto Networks loom, and in scenarios involving Bring Your Own Device (BYOD), contractor access, and similar use cases, Island asserts a unique positioning with limited rival competition. Within the AI spectrum, the startup acknowledges a multitude of emerging startups and established platforms but is keen on consolidating AI controls to streamline implementation for clients, given the impracticality of managing disparate systems.
As they navigate this intricate landscape, Fey indicated that the absence of a cohesive platform would compel businesses to juggle an overwhelming number of AI controls, a situation neither feasible nor affordable in the current economic climate.
Through this whirlwind of developments, Island’s ambition to revolutionize cybersecurity by extending its reach into the domain of agentic workers epitomizes a significant turning point in how organizations adapt to the challenges posed by the rapid advancement of AI technologies.