Pomegra
GCHQ-linked Prevalent AI has secured £16M from a US growth equity firm - its first outside capital in nine years - to push its data fabric platform beyond cybersecurity into broader enterprise risk.
- Prevalent AI raised $22M (£16M) from Los Angeles-based Integrity Growth Partners on August 19, 2026, its first primary capital since founding in 2017.
- The platform maps fragmented enterprise data into a sovereign knowledge graph used by banks, telecoms, and insurers with workforces exceeding 100,000.
- Funds will accelerate US expansion and extend the product beyond cybersecurity into wider enterprise risk use cases.
Lead
Prevalent AI, a London-based data intelligence company with roots in UK signals intelligence, closed a $22 million growth round from Integrity Growth Partners on August 19, 2026. The raise - equivalent to roughly £16 million - ends a nine-year stretch of bootstrapped operation for a business that has quietly served some of the world's largest financial institutions, carriers, and insurers. No valuation was disclosed.
From Spook School to Enterprise Software
The company's founding story is central to how it markets itself. Paul Stokes and Arun Raj built Prevalent AI in 2017 after exiting a prior cybersecurity venture. Their early team included Sir Iain Lobban, who directed GCHQ from 2008 to 2014, and Andrew France, who led GCHQ's cyber defence operations before departing to co-found Darktrace. That combination of intelligence-community discipline and enterprise software ambition shaped what Prevalent AI built: a platform designed to treat data governance as an operational security problem, not a compliance formality.
The company's core product is a data fabric that ingests fragmented enterprise data sources and organizes them into a sovereign knowledge graph - a structured map of what exists across an organization, how those assets relate to one another, and where operational gaps remain. The pitch is that both human operators and AI systems need reliable context before they can make trustworthy decisions. Without it, AI agents operating across enterprise environments work from incomplete or conflicting information.
Why Did the Company Wait Nine Years to Raise?
Stokes made a deliberate choice to take no outside capital for nearly a decade. That matters. Nine years of bootstrapping either means the business was generating sufficient cash internally to fund operations - which signals genuine product-market fit in a segment known for high customer acquisition costs - or it means growth was constrained well below its potential. Probably some of both. Prevalent AI has not published revenue figures, so the baseline against which this round should be judged remains opaque.
What the company did accumulate in that period is a reference base of large enterprises. Banks, telecom operators, and insurers with employee counts sometimes north of 100,000 have deployed the platform, which suggests the product has cleared enterprise procurement and security review at organizations where those processes run long.
What Does Integrity Growth Partners Bring?
Integrity Growth Partners, a growth equity firm based in Los Angeles, led the round. Growth equity is typically patient and revenue-weighted; it sits between venture and buyout, and firms in that category usually require demonstrated revenue before writing checks of this size. The choice of a US-based investor is consistent with Prevalent AI's stated ambition to accelerate its American market entry, where comparable platforms have attracted significant enterprise spend in recent years.
No co-investors were disclosed. The absence of named syndicate partners is worth noting - it may mean the round was structured as a sole investment or that other participants are not yet public.
The Agentic AI Angle
Timing is deliberate here. The enterprise AI market has moved rapidly from experimentation to deployment of agentic AI systems - autonomous agents that take action across enterprise tools and data stores without human intervention at each step. Those systems have a data-quality problem. They inherit the fragmented, siloed information structures that large organizations have accumulated over decades, and acting on bad context can produce bad outcomes at machine speed.
Prevalent AI's framing positions its knowledge graph as the connective tissue that agentic AI needs to operate safely inside complex organizations. The cybersecurity use case - where fragmented data creates immediate operational risk, such as undetected lateral movement or unmapped asset exposure - proved the architecture. The expansion play is to sell the same context engine into adjacent risk domains: operational risk, regulatory compliance, third-party risk management. Large enterprise customers that already trust the platform for security decisions are the natural starting point for that expansion.
Skeptical View
The category Prevalent AI occupies has attracted significant capital and competition. Knowledge graph and data fabric platforms have been built, acquired, and abandoned by major technology vendors over the past decade. What the bootstrapped approach demonstrates is unit economics sufficient to sustain the business, but it also means Prevalent AI has reached a competitive moment - US expansion, product extension, leadership build-out - later than some rivals who took institutional capital earlier. The next twelve months will test whether nine years of deliberate pacing produced a durable competitive advantage or simply delayed the harder scaling questions.
Outlook
The £16M round positions Prevalent AI for its first serious push into the US market, with enterprise risk expansion providing a broader surface area than pure cybersecurity. The GCHQ heritage and Darktrace association have worked well as credentialing in security-conscious sectors. Whether the same positioning lands as effectively with risk officers outside the security function - finance, operations, compliance - is the open question. Integrity Growth Partners has taken a bet that it will.
