Regnology, a leader in regulatory, risk, tax and finance reporting technology, published “The Agentic Gap: From Control to Intelligence in Regulatory Reporting”, accompanied by a video foreword from Chief Executive Officer Rob Mackay. The global study sets out where agentic AI can take on meaningful operational load, and gives institutions a practical roadmap from isolated pilots into scaled production.

The research finds an industry already mobilizing. 87% of respondents are exploring, piloting or embedding AI in operations, and only 13% report no current plan. The constraint is a shared one. Embedded, production-level use sits between 8% and 15% at every institution tier, and 16% of respondents overall have reached it, with 89% of financial institutions yet to get there. Resources buy experimentation, but production use is earned on the same terms everywhere. The report calls that solvable distance, between testing a concept and relying on it inside the reporting cycle, the agentic gap. Where agentic AI cannot yet take responsibility, the report finds the binding constraint is usually the data or the governance around it, not the maturity of the technology.

The economic incentive to close it is substantial. Research by Oliver Wyman, commissioned by Regnology, finds regulatory reporting typically absorbs 1% to 3% of total expenditure at the banks profiled. Between 30% and 50% of reporting spend goes on running the process internally, the largest single cost pool and a reflection of how much of the work is still performed by hand. In illustrative Tier 1 case studies, Oliver Wyman estimates roughly 15% to 25% of reporting spend could be addressable by agentic workflows. The estimate is directional.

Rob Mackay, Chief Executive Officer of Regnology, said: “Banks are conservative by nature, and in regulatory reporting they are right to be. The tolerance for error is close to zero, and a manual process may be inefficient, but it is familiar and readily defensible. The harder part is rarely the technology. It is finding people who understand the regulatory logic in depth and can turn it into something a system can apply safely. That combination is scarce, and it is what Regnology brings alongside institutions rather than asking them to assemble it alone.”

Rather than offering abstract benchmarks, the report provides a practical blueprint for execution. It introduces a diagnostic framework mapping the cost of inaction against AI readiness to help institutions prioritize high-impact workflows, baseline existing processes before piloting, and design governance into systems before deployment. Moving from priority to production means matching the authority given to AI to the risk and repeatability of the individual process, ensuring outputs can be traced and reconstructed, and aligning with requirements such as the EU AI Act.