New regulations, proliferating data and the rise of social media are pushing banks to enhance their use of communications surveillance to identify and prevent financial misconduct.
In electronic communications, e-mails and telephone calls are only two pieces of the surveillance puzzle. With new social media platforms launched all the time, the challenge of separating useful data from irrelevant information calls for carefully designed, integrated approaches. Another problem: Communications surveillance systems do not run by themselves and firms are finding it difficult to attract and retain enough people with the right skills to operate such systems.
New technologies―including advanced filtering tools and artificial intelligence―help reviewers sort through massive amounts of data and identify potentially serious banking compliance problems.
Banks should establish a balance between detection and disposition; in other words, they need to combine “push” methodologies for detection and issuance of high-risk alerts with “pull” methodologies for discovery of new threats and disposition of such threats.
Banks and other financial services firms should consider what has been called “integrated surveillance.”
This approach to communications surveillance uses unified information access platforms to integrate large volumes of unstructured information.
It also incorporates semi-structured information such as metadata and structured information into a comprehensive environment.
For effective compliance initiatives, banks should ultimately implement multiple, complementary surveillance capabilities drawn from a range of technologies including lexicons, machine learning, semantic technology, network analytics and relationship mapping. Visualization tools can help connect and support the various capabilities.
To identify and prevent misconduct, banks, capital markets firms and other financial institutions should employ a range of communication surveillance techniques including:
Enriching unstructured data from e-mails, chats, forums, social media and voice calls with structured information such as internal restricted lists, government maintained lists, transactions and activity blogs
Integrating voice surveillance into the overall surveillance model, using lexicons for transcribed voice conversations
Verifying and centralizing data, establishing new rules, executing ad-hoc queries and defining comprehensive analytics against an extensive data repository.
Other elements of integrated surveillance include semantic technology―which can automatically recognize concepts and topics―along with network analytics to apply statistical methods to analyze networks and map relationships. Visualization tools can also help facilitate discovery.
Read the full report for more insights and additional steps toward an integrated communications surveillance program.