The complexity and ever-changing nature of fraud and financial crime patterns requires an ability to quickly and effectively spot and stop incidents.
Financial services organizations are no strangers to fraud and regulatory issues. However, fundamental changes in the industry and the way organizations conduct business have significantly changed the dynamics and the scope of these issues. Specifically, the rate at which malicious acts, money laundering transactions, or sanction violations transpire is now measured in fractions of a second, and the quantity of transactions that must be scrutinized for regulatory compliance has skyrocketed. Addressing these factors requires continuous intelligence (CI) that uses artificial intelligence (AI) to analyze streams of transactional data in real-time and derive actionable information that can be acted on in milliseconds to seconds.
Given this state of the market, several factors are driving financial institutions to CI and AI. First, penalties for regulatory violations are on the rise. For example, penalties from the Office of Foreign Assets Control (OFAC) of the U.S. Department of the Treasury alone were roughly $1.3 billion in 2019. That is more than six times the total of the previous three years.
Second, the imposition of economic sanctions against individuals and states also is on the rise. The U.S. Office of Foreign Assets Control’s Specially Designated Nationals and Blocked Persons List is 1.364 pages long, which is about 1,000 more pages than at the same time last year.
How CI and AI help
Financial organizations face great challenges as part of their mandated compliance and fraud prevention efforts. The complexity and ever-changing nature of fraud and financial crime patterns requires an ability to quickly and effectively spot and stop incidents.
To accomplish this, organizations must find ways to augment and alleviate the time-consuming, labor-intensive, and often inaccurate processes that have been used in the past by their financial crime operations. An IBM paper titled “Fighting Financial Crime with AI” published last year discussed how cognitive solutions are changing the way institutions manage such operations.
The key areas where AI and cognitive solutions are having the greatest impact are transaction monitoring and sanctions screening alert triage, due diligence reviews, payment fraud modeling, and conduct surveillance investigations. In these areas, financial organizations are using AI, machine learning (ML), and robotic processing automation (RPA) to cost-efficiently resolve the technical and process gaps that criminals exploit today. Here is an example from the paper of CI and AI in action:
Faster Enhanced Due Diligence
A large, regional financial institution was having challenges with its enhanced due diligence (EDD) process. The review process was highly manual, took lots of time, and required lots of data entry. Additionally, results were inconsistent from analyst to analyst and had a high number of errors.
Using AI and other technology, the organization automated data gathering and prioritization. Instead of requiring analysts to gather data, the solution automatically started gathering information on the entity once an alert was triggered. The solution aggregated information from structured and unstructured data sources, including sanctions lists, business directories, and search engines. It then ranked and categorized the data based on its relevance and source.
The next step was to use Natural Language Processing (NLP) to understand the context and sentiment of articles and other information related to the entity being reviewed. This information was then prioritized and annotated for analyst review, helping the analyst understand relevant risk more quickly, as well as why each article was chosen.
Lastly, the solution collected the information used to make the customer risk decision into an automated dossier for easier reference during the audit review, which often takes place weeks or months later.
The IBM paper noted the organization realized several business benefits.
The organization was able to conduct reviews 60 percent faster than before. This reduced investigation times from more than 13 minutes to just over five minutes by automating much of the manual search and data entry process.
It realized a 50 percent reduction in rework. This minimized the need to retrace investigation steps by automatically collecting comprehensive information in the investigation dossier.
From a business impact perspective, analyst productivity increased. They were able to complete investigations in a shorter time, allowing them to conduct more investigations in a given time period. The automation of the data collection process was a key factor here.
Investigation outcomes were more consistent. Using NLP, the organizations applied the same logic across any written information source, eliminating subjective interpretation due to varying interpretations.
Technology Works Hand-in-Hand with Analysts
Perhaps the biggest impact of CI and AI in financial regulatory processes is that the technology complements the work of skilled analysts. In most applications, the technology automates chores and speeds analysis, delivering guidance to analysts that can be used in their decision-making processes.
Aiding decision support is, in fact, one of the common benefits of CI in many financial services applications. In the future, this role is likely to expand into the realm of decision automation.