BIS Innovation Hub's 2025 Analytics Challenge
The Transparency Fabric 2.0
Enriching structured entity data with AI-extracted information
The challenge
According to the Bank for International Settlements (BIS), an estimated $3 trillion was subject to money laundering and terrorist financing in 2024. Combatting this development requires innovative and cost-effective methods to monitor and identify illicit transactions.
A key aspect of financial fraud prevention is the identification of company structures, financial networks and sanctioned entities. As this information is rarely available from a single source, collaboration across multiple organizations is essential to build comprehensive entity profiles. To ensure trustworthiness and reliability, a solution must be fully transparent regarding the source, currency and quality of the underlying data.
Introducing the LEI
At the core of the 'Transparency Fabric 2.0' lies the Legal Entity Identifier (LEI), the 20-character global ISO standard 17442 for legal entity identification.
The LEI connects to key reference information that enables clear and unique identification of legal entities participating in financial transactions, including direct and ultimate parent entities, subsidiaries, branches, managed funds, and umbrella structures.
In addition, the LEI is mapped to other essential standards and identifiers, such as the BIC, MIC, ISIN, S&P Global Company ID, and OpenCorporates identifier, which ensures seamless integration across various financial systems.
Our partners
Open Ownership is driving the global shift towards transparency over who owns and controls
companies by working with governments to make high-quality beneficial ownership data available, including
through advocacy for data publishers to use its Beneficial Ownership Data Standard
(BODS).
OpenSanctions builds a global watchlist product using open-source technology. By combining data from over 80 sources (including government-published sanctions lists, criminal watchlists and international databases of PEPs), OpenSanctions creates a rich graph of companies and people that pose a business risk.
Our solution
The Transparency Fabric 2.0 demonstrates the benefits of maintaining a native mapping between GLEIF, Open Ownership and OpenSanctions data to create a comprehensive network of entity information while applying AI to enrich structured data with information extracted from annual reports.
By applying
Large Language Models, we reduce labor-intensive manual report analysis by programmatically determining entity
relationships. To ensure trustworthy results we place great importance on indicating the sources of both
structured, and AI-extracted information. The creation of entity networks similar to the one in our submission
is only possible by embracing a collaborative spirit among data providers and by standardization of reference
data which enables seamless mapping of datasets. Having the freely accessible and globally standardized unique
LEI at the core of our solution allows the addition of additional sources to establish an even larger
network.