The AM AI PEP Dataset

Politically Exposed Person (PEP) screening is essential for maintaining adherence to know-your-customer (KYC) and anti-money laundering/counter-terrorist financing (AML/CFT) regulations, as well as for mitigating risk and fulfilling regulatory requirements. In essence, effective PEP screening requires a risk-based approach, focusing resources on higher-risk individuals based on various factors which can be time-consuming to gather and analyse.

As a response to this challenge and to make the job of our partners and clients easier, Vartion developed its very own PEP data source for the use within Pascal, our complete compliance eco system designed to streamline and enhance KYC and AML processes for businesses. This dataset, called AM AI PEP, leverages artificial intelligence to identify PEPs, as defined by the Financial Action Task Force (FATF), through analysis of publicly available media. This is crucial for addressing data bias in proprietary lists, ensuring a more comprehensive and accurate identification of politically exposed persons.

By placing a risk-based compliance approach at the centre of the development of the PEP data source, we have embraced the perspective of the FATF, which acknowledges that PEPs, owing to their positions and influence, are susceptible to exploitation for purposes such as money laundering, corruption, bribery, and terrorist financing. However, identifying individuals and establishing their connections to such events can be time-consuming and tedious, especially with the growing volume of media data today, much of which lacks a clear connection to risk-relevant events and connections. Therefore, AI AM PEP addresses this by linking identified PEPs to adverse events, aligned with FATF guidelines and ESG principles, resulting in higher quality PEP hits as only risk-relevant individuals are identified. It also maps the network structures of individuals and organisations connected to PEPs, enabling more precise risk assessments. This approach not only enhances compliance but also reduces operational costs while mitigating legal and reputational risks.

"The Financial Conduct Authority in the UK (FCA) expects firms to make use of information that is reasonably available to them in identifying PEPs, family members or known close associates. This could include the following: Public domain information such as websites of parliaments and governments, reliable news sources and work by reputable pressure groups focused on corruption risk such as Transparency International or Global Witness. Firms should use a variety of sources where possible."

Under this premise, Vartion has successfully developed, and a PEP identification and generation pipeline powered by AI, supplemented with rule-based controls, a project recognised for its innovative character and hence receiving support from the Dutch government (MIT) and indirectly from the European Union (EFRO), both aimed at fostering impactful innovation. Furthermore, the pipeline aligns with ISAE 3402 Type 2 standards, reinforcing its reliability and ensuring rigorous controls in the identification and management of data.

Key Features

Global Coverage: The current AM AI PEP data source offers users comprehensive global coverage including over 1.4 million first-line and more than 400 thousand second-line PEPs, ensuring extensive reach when doing PEP compliance checks.
Reliable and Explainable Results: Each PEP profile you are presented with is backed by several credible, verifiable sources that are linked as supporting documents and media, providing clear evidence for assessments as well as transparency when assessing the PEP’s importance to your compliance approach.
Risk-Relevant Insights: PEPs are identified based on comprehensive media data, with articles categorised as adverse or negative presented in the summary of the AM AI PEP hits in Pascal. This approach provides valuable context, enabling compliance officers to identify risks effectively by leveraging the included information about adverse and negative events.
Regular Updates: At Vartion, we guarantee that the AM AI PEP data source consistently provides you with the most current results. We regularly implement updates to process extensive amounts of curated media, thereby ensuring that the dataset accurately represents the changing statuses and risks associated with PEPs.
Network Mapping: Compared to other tools, AM AI PEP also offers detailed mapping of relationships between PEPs, organisations, and their connections.

Benefits

Regulatory Compliance: Due to the alignment with the FATF Recommendations concerning adverse events, you can ensure greater adherence to global standards for identifying and managing risks associated with PEPs.
Enhanced Efficiency: As mentioned earlier, finding and assessing PEPs represents a time intensive process, which can be greatly reduced when using AM AI PEP as it delivers risk-relevant PEP insights, backed by credible sources, minimizing unnecessary manual reviews and false positives.
Actionable and Explainable Insights: Our data set provides you with clear, evidence-driven results linked to supporting documents, enabling informed decision-making and audit readiness.
Comprehensive Risk Coverage: With AM AI PEP we offer a dataset covering detailed relationships between PEPs, organisations, and adverse events, and hence improving the quality of risk assessments.

Specifications

Data Sources: The compilation of data for AM AI PEP utilizes globally accessible and publicly available English media sources.
Risk-Centric Coverage: The PEP profiles align with FATF guidelines, following a 'once-a-PEP, always-a-PEP' approach, with the consideration of removing a main PEP profile only after six months of confirmed deceased status.
Expanded Fields and Insights: AM AI PEP presents enriched PEP profiles with detailed fields, offering greater context, including roles, adverse mentions, and relationships between PEPs.
Data Quality and Refinement: We utilise AI-powered techniques to remove duplicates and ensure clean, accurate results for better decision-making.