heading

Embedding AI into regulatory mapping processes

description
Streamlining a complex mapping exercise across UK and EU legislation for a global bank
image

The challenge

We recently assisted a global bank with mapping the new Public Offers and Admissions to Trading rules under the UK’s Smarter Regulatory Framework to previous UK and EU legislation. This was a highly complex cross-referencing exercise as the new rules introduced significant changes in terminology and structure, requiring contextual interpretation rather than simple keyword matching; several provisions were consolidated or split, creating one-to-many and many-to-one mapping challenges; and additional regulatory guidance and transitional provisions added further layers of nuance beyond the core rules.

The solution

To address these challenges, we delivered a hybrid human / AI solution comprising:

  • Custom web scraping to extract texts from FCA, legislation.gov and legacy EU sources.
  • A state-of-the-art context-enriched RAG (Retrieval-Augmented Generation) pipeline to align old and new rules intelligently, leveraging semantic similarity rather than simple text matching.
  • AI-powered gap analysis, highlighting areas where obligations have changed or disappeared, all validated through legal team oversight to ensure accuracy.

Results and benefits

The outcome gave the bank’s legal team a clear, actionable understanding of regulatory changes, enabling them to assess compliance impacts and update internal frameworks confidently.

By using AI, this exercise was delivered far more quickly and cost-effectively than traditional methods of manual cross-referencing.

Contact