This function of the technology is a priority for institutions that adhere to strict regulations and need to identify, manage, and reduce the risk of fraudulent operations.
Every day, millions of data are generated that may seem irrelevant, but together they are a key resource for detecting suspicious activities. However, processing and analyzing all this information without advanced technology is almost impossible. For this data to have real value in AML, it is essential to structure and contextualize it.
Óscar Barba
Co-founder & CTO of Coinscrap Finance
Using technology to analyze hundreds of millions of documents per month
As I mentioned, unstructured data is a valuable but underutilized resource in the AML field, precisely because it is very difficult to process. On average, an organization that needs to extract insights about imminent dangers would face the monstrous task of analyzing millions of inputs per month.
AI infrastructure was designed for this purpose, transforming that volume of data into manageable information using advanced natural language processing (NLP) and machine learning techniques.
💡 Did you know…
Generative AI offers endless possibilities for banks.
Explore it today!
5 key tasks of anti-money laundering platforms
Document transformation: Any document, regardless of its original format, is converted into standardized text. This allows sources to be processed and analyzed regardless of their origin (a news article, a report, or a social media post).
Data filtering: The system allows analysts to select and filter information according to their specific needs, creating a highly personalized and relevant dataset for the AML strategy objectives.
Access to multiple sources: In addition to having access to social networks and other public data, the system can integrate the client’s proprietary content or any internal source the entity wants to include, thereby expanding data coverage.
Advanced tagging: Using AI, each document is tagged with key elements such as sentiment, entities (companies, people, locations), specific events, and importance. This makes detailed analysis and risk detection easier.
Data presentation: It also handles presenting all the collected information and its analysis in real-time to teams, ensuring that each department can integrate these insights into their systems efficiently.
The value of unstructured data in anti-money laundering
Leveraging the available (but overwhelming) information makes the difference between early detection of suspicious activities and capital loss due to potential violations.
The disorganized nature of this data makes processing and analyzing it without advanced technology impractical. Fortunately, these tools turn unstructured data into insights, relying on natural language processing (NLP) techniques:
Contextual analysis
It can not only identify keywords related to money laundering but also evaluate the context in which they are mentioned, ensuring that the system can distinguish between relevant mentions and noise.
Sentiment indicators
Various sentiment indicators and dozens of fields describing each detected entity are generated, providing a detailed view. This emotional analysis identifies changes in public perception.
Segmentation by time, region, and topic
Banks and insurers can view filtered data based on these aspects, and in some cases, also by topic similarity, focusing on emerging trends and risk areas.
AI applied to AML: The future risk alert system
Imagine, for example, an alert generated from a tweet mentioning a suspicious transfer. With the right infrastructure, the system can not only detect that mention but also assign it a relevance score based on its context.
This way, analysts can focus on the highest-risk mentions, optimizing their time and resources. Let’s now look at the main features of the alert system:
Contextual scoring-based risk alerts
Each mention of potentially dangerous topics is analyzed to assess its relevance. A series of defined parameters, such as origin, language, country, and source type, are used.
Simultaneous search in multiple languages and countries
This means adapting to the global needs of banks and insurers, ensuring that relevant mentions are captured in real-time. Mapping the globe for potential gaps is indispensable today.
Automated learning with human feedback
As agents review alerts and provide feedback, the system learns automatically. This allows it to fine-tune its analysis and improve the accuracy of future alerts.
An opportunity for the future of fraud detection
The regulatory landscape is becoming increasingly demanding, so the ability to react quickly to potential threats is turning into a competitive advantage. Banks and insurers that adopt this technology and process unstructured data to address these challenges will be better prepared to meet regulations. They will be able to reduce risks and build trust with their clients and regulators.
AI is not only a tool for meeting AML obligations, but it has become a transformative element in the global financial landscape. Fully exploiting this technology is helping institutions become much more efficient, proactive, and secure.
We will soon see GenAI dominating the headlines in this area. It is capable of generating synthetic data and hypothetical scenarios that enrich models, anticipating criminal activities. Protecting the financial integrity of clients and society in general is driving innovations in this field.
Technology is once again the great ally of institutions.
About the Autor
Óscar Barba is co-founder and CTO of Coinscrap Finance. He is an expert Scrum Manager with more than 6 years of experience in the collection and semantic analysis of data in the financial sector, classification of bank transactions, deep learning applied to stock market sentiment analysis systems and the measurement of the carbon footprint associated with transactional data.
With extensive experience in the banking and insurance sector, Óscar is finishing his PhD in Information Technology right now. He is an Engineer and Master in Computer Engineering from the University of Vigo and Master in Electronic Commerce from the University of Salamanca. In addition, Scrum Manager and Project Management Certificate from the CNTG, SOA Architecture and Web Services Certificate from the University of Salamanca. He recently obtained the ITIL Fundamentals certification, a recognition of good practices in IT service management.
