Geordie AI, a London-based startup that helps businesses monitor and control AI agents, has raised $30m (£22.3m) in a Series A funding round led by venture capital firm Balderton. The company says the capital will be used to expand its platform, which is designed to give enterprises greater oversight of the AI systems operating within their organisations. The raise marks one of the more substantial early-stage investments in the AI governance space from a UK-based company this year.

The platform works by providing organisations with a real-time picture of which AI agents are running across their systems, what data and functions those agents can access, how they behave under different conditions, and what risks they may introduce into enterprise environments. Alongside this, the company offers Beam, a remediation suite that allows businesses to adjust and constrain how agents behave without disrupting the pace at which they operate. The distinction Geordie AI draws is between passive monitoring and active intervention: the platform does not simply flag problems but provides the means to address them without requiring businesses to slow or suspend the systems involved. By combining these two offerings, the company aims to solve a burgeoning issue as AI agents transition from experimental phases to operational deployment: the frequent lack of transparency businesses face regarding the actual activities of these systems once they are live.

Henry Comfort, co-founder and chief executive, described AI agents as "one of the most important shifts in enterprise operations in decades." Writing on the company's website following the funding announcement, he said the raise would allow Geordie AI to invest in the team and infrastructure that enterprises need to deploy agents safely and reliably. His stated goal is for Geordie AI to become the standard platform through which large organisations understand, secure and govern their agent operations. Central to that goal, he argued, is building the kind of visibility that allows organisations to extend what agents are permitted to do as confidence in them grows, connecting new systems incrementally and moving beyond contained pilots into sustained, large-scale operational use.

The funding arrives at a point when scrutiny of agentic AI is increasing across industry and research communities. Researchers from Microsoft and Imperial College London have argued that supervising AI agents carries hidden demands that are not yet fully understood or accounted for by the organisations adopting them, and that those demands need to be quantified and built into roles and processes before agents take on more consequential work. That analysis points to a wider concern: that the speed at which AI agents are being adopted inside businesses has outpaced the development of adequate frameworks for managing them. Geordie AI is among a small number of companies attempting to fill that gap with dedicated infrastructure rather than leaving governance to individual organisations to construct themselves.

The Series A proceeds will go towards growing the company's engineering and commercial teams and building out the technical infrastructure required to serve large enterprise clients. Comfort has framed the immediate priority as helping organisations that have already run AI agent pilots to move those systems into continuous, reliable operation at scale. That transition, from isolated test to embedded process, is where oversight failures are most likely to emerge. It is also where the commercial case for a platform like Geordie AI's is clearest: once an agent is running continuously inside a business, the cost of not knowing what it is doing, or not being able to correct it quickly, becomes material. The company is betting that enough enterprises have reached that point to sustain a dedicated market for what it is building.

‍