AI agents need control infrastructure
Corentis Shield is an AI checkpoint for regulated workflows: policy-bound, reviewable and evidence-generating before sensitive actions move forward.
Why now
Corentis sits across regulated-service pressure, AI/agentic AI adoption momentum and a growing need for operational AI control.
AI assurance is becoming a major growth market. The UK government estimates the UK AI assurance market was already worth approximately £1.01bn GVA in 2024 and could reach £18.8bn by 2035 if adoption barriers are addressed. Corentis is focused on one of the sharpest emerging needs inside that market: runtime assurance for AI agents before they take regulated actions.
Within financial services, the adjacent AI-in-RegTech market is forecast to grow from $3.51bn in 2026 to $12.33bn by 2030, creating a strong commercial opening for tools that help regulated firms deploy AI safely, visibly and with evidence.
FCA complaints redress
Total redress in FCA complaints data was £283m in 2025 H1.
This was a 20.0% increase on the 2024 H2 figure of £236m.
Financial Conduct Authority, 23 October 2025
McKinsey global AI survey
23% of respondents said their organisations are scaling an agentic AI system somewhere in the enterprise.
A further 39% said they had begun experimenting with AI agents.
McKinsey & Company, 5 November 2025
IBM / Ponemon
63% of breached organisations lacked AI governance policies to manage AI or prevent shadow AI.
IBM also reported that 97% of organisations with an AI-related security incident lacked proper AI access controls.
IBM / Ponemon Institute, 2025
The wedge
Corentis starts with financial-services complaints and vulnerable-customer workflows: narrow enough to pilot, concrete enough to evidence, and sensitive enough to prove why checkpointing matters.
The aim is to prove the checkpoint/control pattern in a high-consequence workflow, then expand into adjacent regulated workflows where AI outputs affect customers, records, decisions or operational actions.
The platform logic
Why this is not just another AI tool
Corentis is not a chatbot, model wrapper or dashboard after the event. The product sits at the action boundary where AI output becomes operational behaviour.
The company is early-stage and developing pilot-ready prototypes. The current aim is to prove focused design partner workflows before broader commercial rollout.
ControlBench UK as a strategic asset
ControlBench UK is being developed as a strategic AI assurance dataset and evaluation harness for regulated AI agents. It supports the wider infrastructure story behind Corentis Shield.
The asset brings together regulated-workflow scenarios, vulnerability-handling tests, policy-control checks, human-review gates, blocked-action examples, escalation logic and audit-ready evidence outputs.
Corentis Shield is the commercial checkpoint layer. ControlBench UK helps validate the scenarios, controls and evidence model that make the checkpoint credible.
What Corentis is seeking
Strategic support and validation use cases
Corentis is preparing an early-stage funding round and is interested in speaking with experienced angels, strategic partners and investors focused on AI assurance, RegTech and regulated AI infrastructure. This page is for interest and conversation only; it does not offer shares or investment terms.
Product development
Strengthen Corentis Shield as the commercial runtime-control layer.
ControlBench UK
Develop the strategic AI assurance dataset and evaluation harness.
Sandbox and pilot validation
Prepare financial-services complaints and vulnerable-customer workflow testing.
Design-partner work
Turn real workflow insight into scenario libraries, controls and evidence outputs.
Vision & Funding Readiness Overview PDF
A public-facing overview of the Corentis vision, the timing, the first wedge and the validation path for serious funding conversations.
Best for funders and investors who want to understand why now is the moment to build AI checkpoint infrastructure.
ControlBench Innovation Brief PDF
An innovation-led brief explaining how Corentis can benchmark the difference between uncheckpointed and checkpointed AI-assisted workflows.
Best for readers interested in the technical validation path behind Corentis Shield.
UK Strategic AI Infrastructure Memo PDF
A bold but careful memo on the UK opportunity for AI control infrastructure between agents and sensitive regulated actions.
Best for strategic funders and partners considering the next layer of responsible AI infrastructure.