The benefits of adverse media screening for banks and financial institutions

Why is Negative News a hot topic in 2023? Ripjar’s Gabriel Hopkins shares more about the importance and benefits of adverse media screening for banks and financial institutions.

Ripjar helps banks, governments, and other organisations monitor, detect, and tackle different forms of criminal activity using sophisticated technological solutions including AI, Machine Learning, Natural Language Processing (NLP), and graph analytics.

Founded by 5 experienced data analysts and technologists from the UK’s Government Communications Headquarters (GCHQ), Ripjar specialises in making sense of vast quantities of structured and unstructured data.

Banks, Financial Institutions, payments companies, and corporates in general face significant challenges today in successfully screening customers and other counterparties to understand risk – particularly money laundering and terrorist financing.

Ripjar’s Labyrinth Screening solution takes an innovative approach to matching counterparties to different registers of risk – including PEP (Politically Exposed Persons), sanctions, and watchlists as well as millions of media articles that highlight important risks – referred to as Adverse Media or Negative News.

Adverse media screening – often referred to as Negative News – is an important subset of customer due diligence (CDD). Financial Services regulators around the world are recognising that the ability to identify risks from allegations and other reporting in the media can play an important role in tackling money laundering and terrorist financing. In other sectors, companies are utilising adverse media screening to understand additional signals such as Environmental, Social, or Governance (ESG) risks or even news of data compromise.

Successfully identifying pertinent data from the millions of unstructured news articles published each day is a complex challenge. Ripjar’s Labyrinth Screening uses AI and Machine Learning to essentially read through each article, filtering out relevant content and identifying the relevant facts – who did what to who, where, and when.

In 2022, Ripjar introduced AI Risk Profiles. By understanding common factors related to the people and companies in each article, the system is able to construct condensed profiles that capture the essential risks related to each entity, reducing the complexity of reviews for compliance officers by as much as 90%.

The final stage is to match a bank’s customers (or other counterparties) against the adverse media. Optimised name matching is another important element. It is important to be able to match name variants – such as Jon, John, and Johnathan – as well as more complex misspellings and script variants.

Banks and other organisations are exposed to counterparty risks on an ongoing basis. They need to take a careful risk-based approach to ensure they are meeting often-onerous regulatory obligations without impacting their business.

Regulators around the world continue to increase expectations around customer due diligence. The most recent 6th Anti-Money Laundering Directive (6AMLD) in Europe requires regulated entities to put systematic Adverse Media checks in place alongside PEP, sanctions, and other watchlist checks.

With enhanced Adverse Media capabilities, banks will be able to identify different types of risk that may not be covered by traditional sanctions watchlists. It is important with such a powerful early warning system to respond quickly. With the latest generation of screening, monitoring is continuous, with alerts generated and returned to the bank for assessment by human analysts to determine their significance or severity as soon as they emerge.

The proliferation of so-called fake news and targeted misinformation online provides a significant challenge when implementing a balanced system of checks. It is essential to incorporate trusted data sources such as premium news providers. Missing significant updates or regulations can result in violations and potential fines, so it is important to test and validate the solution on an ongoing basis.

Another problem with unstructured data sources is making incorrect inferences. Many unsophisticated systems are driven by the proximity of names to risk keywords. Unfortunately, such an approach disregards the complexities of different languages. A common problem is identifying judges, journalists, or even victims as the perpetrators of different crimes.

By using Natural Language Processing (NLP) and machine learning, it is possible to interpret these mistakes and avoid unwanted associations.

The UK and countries in the European Union are required to implement all European Anti-Money Laundering Directives into local law in a timely manner. The EU has required banks and other regulated firms to consider Adverse Media controls since the 4th Anti-Money Laundering Directive, and the 6th Anti-Money Laundering Directive (6AMLD) tightens controls further, requiring a systematic Adverse Media process.

Different regulators have different approaches, but around the world, there is a consensus that Adverse Media provides a powerful approach to risk identification. Regulators such as MAS in Singapore and HKMA in Hong Kong, amongst others, have led the way in requesting prudent Adverse Media controls are put in place.

Ripjar’s AI Risk Profile technology represents a substantial step forward in the identification and management of risk data, in a high-quality screening solution. By identifying commonalities between different people and companies in the data, it is possible to vastly simplify the effort required to screen customer portfolios.

As banks and other organisations look to utilise broader screening checks to lower-risk and retail customers, AI-driven models can analyse millions of data points daily to optimise matches while reducing both false positives and false negatives.

Current generative AI techniques tend to be very good at summarising different data and activities that they have access to, but are limited in terms of up-to-date data sources, as model training is performed on historic data sets, meaning they will lack relevancy when looking for emerging risks. Current feedback is also that these models can be badly wrong and it is very difficult to calibrate accuracy without fully reviewing the source materials.

However, the technology is clearly advancing quickly and Ripjar is researching how generative AI and specifically Large Language Models (LLMs) can be used to detect risks in broad sets with even greater levels of accuracy.

2022 was a fantastic year for Ripjar as we introduced multiple new innovations - such as AI Risk Profiles - and expanded the range of customers we work with. We were excited to announce an enhanced partnership with Dow Jones Risk & Compliance at the start of the year and we will continue to innovate to find better ways to help banks and other organisations detect and tackle criminal activity.

About Gabriel Hopkins

As Chief Product Officer, Gabriel leads Ripjar's product management with the responsibility for setting and delivering the company's product strategy. With 20 years of product management and marketing experience, Gabriel understands the challenges of developing sophisticated software solutions while balancing innovation with customer needs. Prior to Ripjar, Gabriel was Vice President, Product Management at FICO overseeing the development of the company's next-generation fraud and financial crime solution. He has also served as Vice President, Product Management and Marketing at WorldPay (now FIS).

About Ripjar

Ripjar is a global company of talented technologists, data scientists, and analysts designing products that change the way criminal activities are detected and prevented. Ripjar’s founders are experienced technologists & leaders from the heart of the UK security and intelligence community all previously working at the British Government Communications Headquarters (GCHQ). Ripjar uses artificial intelligence and machine learning to augment analysis that can protect global companies and governments from the ever-present threat of money laundering, fraud, cyber-crime, and terrorism.