Advancements in RegTech have been impacting business and process innovation
Gabriel Chan, Head of Global IT, Gaw Capital
Gabriel Chan, Head of Global IT, Gaw Capital
Transaction Monitoring
Transaction monitoring is a key function of AML is the reporting of suspicious transactions, mainly based on some static criteria such as sanction list, payment threshold, and keywords filtering. And is a labour-intensive operation of AML compliance. RegTech solution makes use of Machine Learning (ML) technology to perform big data analysis on all transactional information and customer behaviour. It can more accurately flag those suspicious transactions with less false positive, also helps effectively categorize and rate the unknown activities to prioritize for further investigation.
Internal Surveillance Monitoring
Financial Institute requires to ensure the business process or individual practice comply with the company policy and regulations, such as market abuse, insider trading, conflicts of interest and other non-compliance violations. The internal control function was usually monitoring by keyword search and static filtering rules to detect any potential violation from different channels including emails and instant messaging. Since the communication data is usually in un-structural format, which would be even more complicated than transaction monitoring. By using AI to perform comprehensive big data correlation, RegTech solution can significantly increase the monitoring effectiveness.
Voice Recording Review
Another challenge in compliance monitoring is the voice communication review, for example, in the old fashion trade confirmation voice call recording, or FinTech was using voice as a user interface for client verification. Natural Language Processing (NLP) technology can help accurately recognize and convert the speech into a text script. The automatically generated searchable context would enable for further control monitoring and investigation analysis. In the old days, the control effectiveness varies as highly depends on the unmanageable human factors listening to the voice records word by word. With RegTech, the computer-generated result should be more consistence and is possible for automated process integration. Additionally, AI could provide value-added insight from the conversation such as recognizing the speech emotion. Just be aware, the conversion accuracy for non-English or local languages could be lower as it will depend on the technical design and its data model sampling size.
Expense or Investment?
Traditional institutes with limited resources may have to perform the monitoring on risk-based approach or sampling basis. However, the manual sampling approach is not efficient nor effective, and is not dynamically scalable to cope with the fast business change.
People worry about the AI machine may take over the human job as in the movie? If I use the experience in cyber security, our team was in fact developed a lot of automation tools to assist ourselves in managing the huge amount of security events. Hence, we can free up the limited manpower to perform the more advanced forensic investigation case.
I believe it shall be similar with RegTech, the pressure on the team is always high as to adhere to global regulatory demands. With the advancements in RegTech, it could filter and extract the real valuable insights, helping to unleash the collective power in your well-trained professional people. The investment eventually creates a more nimble and scalable control process to help the company comply with the increasingly stringent and ever-evolving regulatory changes.
See More: Top RegTech Companies in APAC
