Financial regulation has always required comparison, but the tools to do it well have never kept pace with the complexity of modern markets. As supervisory teams face mounting volumes of data, fragmented documentation, and increasingly cross-border policy challenges, artificial intelligence is emerging not as a replacement for regulatory judgement, but as the analytical foundation that makes that judgement sharper, faster, and more defensible.
Sherlocq, a purpose-built RegTech platform shaped in part by former senior figures at the Dubai Financial Services Authority (DFSA), is positioning itself at the centre of this shift.
Sherlocq recently discussed regulatory intelligence and the debate around rethinking oversight in the age of AI.
The core problem is one of scale. A single thematic review can now span hundreds of firms. Policy consultations generate thousands of responses. Supervisory teams are expected to draw systemic conclusions from heterogeneous, fragmented data, yet the manual processes that underpin most regulatory analysis have changed little in decades. The gap between what regulators need to know and what they can efficiently process is widening, and it is widening quickly.
Regulatory benchmarking is more nuanced than simple data comparison. It operates across at least three dimensions simultaneously. Structural benchmarking asks whether a firm’s governance arrangements, policies, and frameworks are substantively comparable to those of peers with similar size, business model, and risk profile, a task that demands document interpretation, not just data aggregation. Behavioural benchmarking examines observable conduct, from complaint handling to trading patterns, asking whether a firm falls within the normal distribution for its sector. Policy benchmarking, meanwhile, assesses proposed or existing frameworks against international standards: what the Monetary Authority of Singapore (MAS) has implemented on digital assets, what the Financial Conduct Authority (FCA) has done on consumer duty, what the Abu Dhabi Global Market (ADGM) has introduced on sustainable finance disclosures.
Each dimension is, by traditional methods, enormously time-consuming. Sherlocq was designed to change that calculus. The platform’s analytical architecture is grounded in a straightforward but powerful insight: the most valuable regulatory intelligence already exists. It is locked inside documents, policy frameworks, examination reports, consultation responses, and firm submissions. The challenge is extracting, structuring, and comparing it at scale.
In practice, Sherlocq allows supervisory teams to ingest large volumes of regulatory documentation and apply structured analytical queries across the entire corpus. A team conducting a thematic review on operational resilience can use the platform to extract how each in-scope firm describes its critical business services, recovery time objectives, and scenario testing methodology, then generate a comparative matrix that surfaces outliers, gaps, and emerging patterns.
Work that would previously require weeks of manual analysis by multiple analysts can be completed in hours. More importantly, the comparison is exhaustive rather than sampled: every document is read, every relevant passage identified, every response coded against the same taxonomy. When every document is read and every response coded consistently, supervisory consistency stops being an aspiration and becomes an operational standard.
The platform also addresses a growing need for cross-jurisdictional policy analysis. As bodies like the DFSA, ADGM, MAS, and FCA design frameworks for rapidly evolving areas including digital assets, AI governance, and sustainable finance, the traditional approach of manually reading a handful of peer frameworks and synthesising observations in a memo is no longer sufficient. Sherlocq enables regulators to upload a curated corpus of international frameworks, consultation papers, and technical standards, including guidance from the International Organisation of Securities Commissions (IOSCO) and Financial Stability Board (FSB) recommendations, and query that corpus systematically. A policy team comparing digital asset disclosure requirements across IOSCO, MAS, and ADGM can run that analysis in a single session, producing structured, sourced output in minutes rather than days.
Accountability is a legitimate concern when AI enters the supervisory workflow. Sherlocq addresses this directly: every analytical output is traceable to its source material, grounded in explicit textual evidence rather than opaque model inference. This also addresses the consistency problem that plagues manual supervision, where different examiners assessing similar firms using different internal benchmarks can produce divergent outcomes that are difficult to justify. Methodological standardisation across a peer group improves the defensibility of regulatory judgements without removing human decision-makers from the process.
Certified to ISO 27001 and ISO 27701, with audit trails built into every output, the platform also incorporates sanctions intelligence that queries more than 320 data sources spanning global regimes including OFAC, OFSI, EU, UN, and UAE designations in a single search. Sherlocq describes this as making it the first AI-native platform to deliver this depth and traceability across multiple sanctions regimes simultaneously.
The question for forward-looking regulatory bodies is no longer whether AI belongs in the supervisory toolkit. It is whether they are investing now in the data infrastructure and analytical tooling that intelligence-led supervision demands.
That means choosing purpose-designed platforms that understand the nature of regulatory documents, the importance of source attribution, and the non-negotiable requirements of explainability and auditability, not generic enterprise AI repurposed for supervisory use. For teams at the DFSA, ADGM, MAS, FCA, and beyond, the infrastructure to do that well already exists. The question is how quickly they move to deploy it.
Bhavin Shah, CEO of Sherlocq, said, “Every regulatory body faces the same problem: the volume of data requiring analysis has outpaced the capacity of traditional supervision. Sherlocq turns that fragmented data into structured intelligence, so that a compact team at the DFSA or MAS can achieve the analytical reach of a much larger organisation. Regulatory AI is not coming. It is here.”
By Daniel Willis, Editor of RegTech Analyst
Read the full Sherlocq post here.
Copyright © 2026 RegTech Analyst
Copyright © 2018 RegTech Analyst
