June 26, 2017
Beng Ti Tan is Head of Compliance, Asia at Fujitsu, based in Singapore. Nicholas Turner is a lawyer with Clifford Chance in Hong Kong, where he is a member of the firm's economic sanctions and financial crimes practice. The views presented are the authors' and do not constitute legal advice, nor do they represent the views of the authors' respective employers.
It's a losing battle. Compliance professionals are facing off against a mountain of information as they try to protect their companies from third party risks. Whether it's hunting down Politically Exposed Persons (PEPs), pouring over matches to the Office of Foreign Assets Control (OFAC) sanctions lists, or performing routine due diligence checks, human operators are spending countless hours searching, comparing, and verifying amid a rapidly expanding world of data against a backdrop of intensifying regulatory expectations. The typical approach—which entails performing sequential searches and manually reviewing possible hits or red flags for compliance risks—is sub optimal to say the least. Search results are dominated by false hits, while the number of data points, transactions, customers, sanctions targets, negative news, PEPs, and other inputs are increasing every day.
At the current pace, the data will eventually overwhelm our screening systems. Errors will rise and with them violations and penalties. Already companies have paid record fines as a result of KYC and name screening deficiencies. hundreds of millions of dollars is considered normal. Worse still, customers must endure their own costs in time, money, and frustration from inefficient compliance measures.
The time is now for a fresh look at how we've designed our KYC and name screening tools. New technologies are offering new opportunities, such as:
- Semantic and relational searching: instead of searching sequentially through data, semantic mapping can find dynamic relationships in data, offering quicker, more intelligent results.
- Big Data processing: the world of information expands in limitless fashion, while legacy systems apply outdated coding and logic to fragmented data sets that are slowing down day by day. Big Data solutions offer better ways to store, retrieve, and search data.
- Intelligent Inputs: KYC and name screening should draw on historical data that may not be in electronic formats. New input methods can intelligently read documents and extract information for integration into the Big Data pool.
- Predictive algorithms: machine learning applied to KYC and name screening results can weed out false hits and instantly target similar profiles for additional assessment, thereby leveraging the value of historical hits in a system that grows smarter over time.
- Ultimately such systems need to be integrated into auto-resolving blockchain type transaction systems as we transition to a future economy.
Like most, we're not technology gurus, but we are keen to explore these future systems—and say goodbye to the outmoded systems of today. Curious? Read more about it in our discussion paper available here.
Readers from all companies (and from the TRACE community) can share their thoughts anonymously in a short survey on this topic. The results will be published in Fall 2017 on the TRACE blog with your observations on how future KYC and name screening systems can be designed and implemented today. Thanks in advance for taking the survey!