Artificial Intelligence, Internet of things, robotics – we’ve been hearing about the vast and disruptive ramifications of these innovations for the past few years but what do they mean for trade finance, particularly in the efforts to combat trade-based money laundering? 

The Fourth Industrial Revolution 

The Fourth Industrial Revolution, often referred to as 4IR, is characterized by a range of new technologies (such as those mentioned above) that are fusing the physical, digital, and biological worlds, impacting all disciplines, economies, and industries.

The term was coined by Professor Klaus Schwab, Executive Chairman of the World Economic Forum and it was also the main theme at the Davos conference back in 2016. Fast forward to 2021 and we can say that the industry has felt the disruption caused by 4IR. Smart technology has changed how we live and work, and it is not going away any time soon. 

The acceleration of the 4IR 

Financial institutions began down the path of digitization long ago, however, pandemic-induced shutdowns caused 4IR to accelerate at a rapid pace. This acceleration is known as disruption: a drastic, sweeping change to industries or businesses because of technological advancement. 

McKinsey & Company shared a series of trends that will define 2021. One such trend is digitally enabled productivity that speeds up the fourth industrial revolution. It is interesting to note that, the pandemic has increased the speed of digitization by 3-7 years and that since January 2020, productivity has improved by 4.1%.

Digital city

Digitization in trade finance compliance 

There are several smart technologies that have been incorporated into the trade finance industry so far. 

Optical Character Recognition (OCR) 

To put it simply, Optical Character Recognition (OCR) is the technology that can turn an image of text into a digitally editable format. This is especially useful for trade finance as a business dealing with hundreds of documents. OCR is a tool that fills a need for compliance review and party screening.

A manual name screening process is incredibly tedious and exposes the business to risks. For instance, manual processes are prone to human oversight or spelling errors that can affect matching against watch lists. 

Through digitalization and the use of OCR, data entry for name screening can be automated resulting in higher accuracy rates. This leads to a more efficient process and enables operations to perform stronger risk analyses and make sound decisions. 

Machine Learning (ML) 

The implementation of Machine Learning (ML) is another step in the digitalization journey of trade compliance. Through predictive analysis, data can be examined with ease. This is not to say that all analysis is left to the system, but rather that the process is optimized so that less time is spent on tasks that can be automated.  

The process of teaching the system how to recognize templates, group data, and set corresponding values is known as supervised machine learning. 

After mastering supervised ML, businesses can look into deep learning which is ML on steroids: it uses a technique that gives machines an enhanced ability to find and amplify even minute patterns. 

Understanding the data 

There is intelligence to be gleaned on the data captured from OCR and ML. This is where the industry is heading next.

The true power of ML lies in deciphering the underlying trends on the data. By creating an active repository of details accumulated from OCR we can formulate a preventive approach to combating fraud and money laundering efforts. 


How would this work in theory? 

Hundreds of transactions pass through operations daily, ensuring that no sanctioned or fraudulent entities are involved seems like an impossible task.  

As we build databases through OCR and ML, we strengthen the predictive analysis capability of our systems. Deep learning would then fill the gaps and provide a second layer of control when we are conducting compliance checks and performing name screening.

Human memory is imperfect, and we couldn’t possibly remember every sanctioned entity. Having an intelligent system that would warn us whenever it detects an unwanted party can protect us when human recollection fails. 

Furthermore, the predictive analysis would also help in terms of customer activity monitoring. Through the data that is gathered from OCR and Machine Learning, we get a stronger understanding of the usual behaviour of clients. This technology can safeguard against fraud.  For example, when a transaction is out of character for the client, we would be warned to check if the details of the request are valid. 

The need for compliance

4IR technologies are being implemented at an unprecedented rate. Financial institutions are on pace to deploy more sophisticated Deep Learning and Advanced Analytics tools in their processes. 

Good progress has been made across the industry, however, more can be done to revolutionize the way it handles AML Compliance. We have a long way to go to realize the full benefit of smart technology in the trade value chain. 

As part of those steps, the combination of OCR and ML paves the way for trade finance departments to use digitalization in a way that improves efficiency and strengthens controls. When machines learn, operations save time and reduce the risks of oversight. For compliance, the greatest potential benefit comes from predictive analysis which allows for the anticipation of fraud and red flags. 

Broadly, the application of AI to AML/CFT processes enhances the capabilities of actors to respond to risks and implement requirements more effectively. These tools are not a replacement, but rather a complement to the systems aimed at improving results and simplifying compliance. Transaction monitoring using AI and machine learning tools may allow regulated entities to carry out traditional functions with greater speed, accuracy, and efficiency. 

The challenge now is for more institutions to use this technology and share their learnings both with regulators and their peers. Internally within the business, ambassadors should encourage the transition from manual data entry to AI. This would not replace the important human analysis component needed to facilitate each transaction. Rather, adapting to this new way of working will further enhance the controls that are already in place. As with any change, starting is difficult but the returns are well worth it. 

As someone who uses OCR and ML on a day-to-day basis, I can say that it has tremendously improved the way we conduct our compliance checks. Operations are transitioning to the digital way of working and people are keener to use the tools available to them. We have realized efficiency gains and strengthened controls by integrating smart technology into our processes. It is amazing to see how we are adapting to this disruption and exciting to know where it will lead us in the coming years.