Regulators can interrogate reported data faster and in greater detail. Compliance controls will need to keep pace.
That pressure extends beyond regulatory reporting. Artificial intelligence (AI) is moving into surveillance, regulatory change and financial-crime workflows. Firms must now demonstrate that the data supporting those systems is accurate, consistently represented and traceable to its source.Speakers at A-Team Group’s RegTech Summit London on 1 October describe a common requirement. Firms need to move from correcting problems after submission towards controls that can show, continuously, why their data and automated decisions should be trusted.
The summit, now in its tenth year, will bring together regulators, financial institutions and technology providers at etc.venues, 8 Fenchurch Place, London.
AI Authority Brings Accountability into the Workflow
The opening AI panel will move the discussion from data foundations to automated action. Panellists will examine how firms are taking AI into production, how agentic systems change decision-making and where human approval must remain.
Fahreen Kurji, Chief Customer Intelligence Officer at Behavox, will join compliance leaders from Rabobank and BTIG, Chrissy Hill of Dreadnought Solutions and Vall Herard of Saifr.
The governance question becomes more difficult as an AI system gains access to more data, tools and workflows. Firms must define which actions it can take, the evidence it must retain and who is accountable when its conclusion is wrong.
Supervision Moves Beyond Periodic Remediation
Supervisors increasingly use analytics to identify anomalies, missing data and weaknesses across submissions. That shortens the time firms have to find and correct problems before regulators do.
“Regulatory reporting is moving into an era where supervisors can interrogate data with a level of speed and sophistication that simply wasn’t possible before,” said Andrew Pinnington, Chief Customer Officer at Novatus. “Firms therefore need to move beyond periodic remediation and build controls that continuously demonstrate the accuracy, completeness and timeliness of what they report.”
Pinnington will develop this theme in his keynote on accurate, complete and timely reporting. The session will consider how firms can improve reporting quality while reducing the cost of repeated investigation and remediation.
Continuous control does not necessarily mean every process must operate in real time. It does require firms to identify errors earlier, preserve evidence and understand how data changes between source systems and regulatory submissions.That places greater weight on lineage, reconciliation and ownership. A control that flags a problem without showing its origin may accelerate detection while leaving the investigation largely unchanged.
AI Cannot Repair Weak Data Foundations
AI can examine larger datasets and find relationships that fixed validation rules may miss. Its output remains dependent on the records, identifiers and definitions supplied to it.
“AI has enormous potential to transform data quality, but it cannot compensate for weak foundations,” said Stephanie Rubizhevsky, Chief Solutions Officer at Novatus. “The real opportunity comes from combining intelligent automation with consistent standards, strong governance and transparent controls to create data that firms and their regulators can genuinely trust.”
Tom Soden from Novatus will join Dawd Haque of Deutsche Bank, Hussain Abdullah of Citigroup and Dan Matthews of MarketAxess for the panel on unifying data, standards and AI.
Their discussion will examine whether AI-driven data-quality tools can move beyond rule validation across fragmented systems. It will also consider the role of the Common Domain Model (CDM), Unique Product Identifier (UPI) and Legal Entity Identifier (LEI).
Standards provide a common representation, but adoption alone does not resolve differences between systems. Firms still need to map legacy data, govern changes and establish which version of a record supports each obligation.
Building Data That Can Be Reused
Regulatory reporting programmes have often developed separately, even when they draw on the same trades, counterparties and products. Each implementation can introduce another transformation layer, reconciliation process and set of controls.
“As regulatory requirements and market transparency obligations continue to evolve, the industry’s biggest challenge is no longer producing reporting data, but creating a consistent, trusted representation of data that can be reused across multiple obligations,” said Dan Mathews, Senior Product Manager of Reporting Services at MarketAxess.
“By combining common standards, robust governance together with AI-driven data quality, firms can reduce complexity, improve transparency and move closer to a ‘report once, use many times’ model. I look forward to discussing these challenges and opportunities at the RegTech Summit London.”
A reusable representation could reduce duplicated mappings and help firms apply corrections consistently. It also demands agreement about ownership, definitions and the controls required when the same data supports several submissions.
The summit will examine the architectural choices behind that model. These include centralised repositories, virtualised sources of truth, application programming interfaces and managed reporting services.
From Conference Discussion to Practical Comparison
The afternoon programme will apply these questions to unstructured regulatory information, trade surveillance and financial crime. Champagne Roundtables will give practitioners an opportunity to compare approaches to reporting data, perpetual KYC controls, AI governance and human judgement.
- Can firms trace an automated conclusion to reliable data?
- Can they explain why the system acted or remained silent?
- Can the same controlled information support more than one obligation?
These discussions should provide a practical test for emerging technology.
RegTech Summit London takes place on 1 October 2026 at etc.venues, 8 Fenchurch Place, London. View the full agenda and register HERE.
Subscribe to our newsletter
