6AMLD & FATF & Adverse Media Screening
In the EU, the 6th Anti Money Laundering Directive (6AMLD) compels firms to perform enhanced due diligence processes for high-risk customers, including
- “Carrying out open source or adverse media searches.”
A thorough AML programme has many moving parts, and adverse media screening is arguably one of the most important for preventing financial crime.
Onboarding a client that has known risk
- Could have severe consequences for your reputation, and
- Also puts you in the firing line if you inadvertently aid activities like money laundering, fraud or terrorism financing.
According to the 2020 Anti-Money Laundering Preparedness Survey Report by Deloitte, however, only 63 per cent of respondents indicated that they undertake regular adverse media searches to update customer profiles.
This suggests that there’s a general lack of awareness among firms of the importance of adverse media checks – and financial institutions should be aware of the relevant regulatory framework where they operate and ensure that they are meeting compliance standards at all times.
What is adverse media?
Adverse media refers to relevant negative news or media coverage about a potential customer or client found across a wide range of sources.
Thorough adverse media screening is useful in exposing a person or organisation’s involvement in activities like money laundering, financial fraud, terror financing, and organised crime, especially when they are targeted at higher risk crimes.
Adverse media scanning transcends all the different types of media sources, from traditional news sources to blogs, web articles, and even online databases.
EU regulatory requirements
In the EU, the 6th Anti Money Laundering Directive (6AMLD), which comes into force from June 3, 2021, compels firms to perform enhanced due diligence processes for high-risk customers. This includes
- “Carrying out open source or adverse media searches” and
- Encourages the use of automated adverse media screening to do this.
- Clients could be automatically categorised as high risk based on factors such as where they are located.
6AMLD replaces the fifth version of the Directive and extends it to add both cybercrime and environmental crime to the list of designated offences linked to money laundering and terrorist financing.
It also increases the criminal liability of money laundering to “enablers” – in other words, financial institutions that fall foul of meeting their obligations under AML/KYC regulations.
EU Member States are required to implement the 6AMLD while keeping their AML/CFT (‘Combating the Finance of Terrorism’) laws aligned with the recommendations of the intergovernmental Financial Action Task Force (FATF).
FATF guidelines recommend that adverse media searches should be performed as part of an enhanced due diligence process and that where a customer has been negatively mentioned in the news, this could indicate a higher risk that requires extra precautions. The guidelines provide clarity on which offences firms need to be aware of and refer to these as the “designated offences” – e.g., fraud, counterfeiting, piracy, smuggling, extortion.
Approaches to adverse media categorisation
The alignment of 6AMLD and FATF illustrates the importance for financial institutions to ensure that they’re taking the right approach to searching for and categorising adverse media.
Failure to do so could lead to relevant adverse media slipping through the net and opening up financial institutions to liability for their clients’ crimes, especially if they have previously faced regulatory penalties or investigation for money laundering, terrorist financing, or other relevant activity.
General news categorisation
When categorisation is general news topics, it loses its specific AML/CFT risk-based focus.
This approach can see firms failing to meet requirements under 6AMLD/FATF because they’re not able to adequately choose categories according to relevant risks.
It increases the chance of missing something important and exposes the firm to a higher risk of allowing suspicious activity or information relating to designated offences to slip through the net and go unnoticed.
A general news classification is usually associated with tools that are designed for media monitoring for non-compliance use cases.
Not only do these tools lack an appropriate categorisation, but they also make screening unnecessarily inefficient by returning high amounts of individual news articles and not clear profiles of people & organisations.
Another common approach is to categorise adverse media based on a set of keywords, i.e., by inputting certain keywords into a search engine and seeing what it spits out.
Keywords are fragile, however, and the dynamic nature of search engines (e.g. Google) can reduce the efficacy of this approach.
For it to be effective, you need to
- Include all variations and synonyms of the word you’re searching for: fraud, fraudulent, fraudulently, fraudulence.
- Consider including and omitting modifiers such as “and” and “or”, as these can affect the search result, and the limitations of only conducting searches in English.
Even with a thorough set of keywords, search engine results are far too lengthy to be useful.
Even with the sheer number of results delivered for a given search query, you often wouldn’t come close to scratching the surface of what’s actually out there, as many results will be false positives or omitted entirely because they aren’t contained within currently indexed web content.
Trawling through these results wastes the valuable time of analysts that could be better spent elsewhere.
Automated machine learning categorisation
Manual adverse media screening is time-consuming and ineffective; it’s all too easy to miss important, relevant information and fall short of regulatory requirements due to the inherent limitations.
Automated systems on the other hand, such as those powered by now established (but still rapidly improving) technologies like artificial intelligence (AI) and machine learning (ML), optimise the screening process.
In addition to saving time and cutting costs, ML-driven automated screening helps financial institutions mitigate risk and conduct thorough, diligent investigations into potential clients.
With natural language processing, they’re also able to ignore irrelevant news and flag results that may require extra scrutiny from a human analyst.
Systems can easily be configured to carry out thorough searches and conduct efficient daily ongoing monitoring of both current and potential clients, and any information that is discovered will automatically be categorised according to its context in order to help drive relevant & timely alerts.