Many financial services organizations are attempting to address this challenge through a broad, forward-looking set of analytically driven technologies. These are collectively referred to as RegTech, or individually by their underlying artificial intelligence or machine learning-driven (AI/ML) technologies. In fact, a recent FINRA report on the adoption of AI in the securities industry indicated that 77% of all financial services respondents anticipate AI to possess high or very high overall importance to their businesses in the next two years.

Both the SEC and FINRA are aggressive consumers of AI/ML in their inspections tools and have stated publicly that they expect firms to be embracing these technologies in their compliance programs.

Unfortunately, the discussion of RegTech or underlying AI is often oversimplified as an either/or battle of human versus machine. One extreme argues that we’ve finally arrived at the tipping point where the keyword-based inspection of information is now producing diminishing returns. The opposite perspective stresses that AI requires a significant level of investment and training to develop even a rudimentary understanding of a firm’s regulatory policies given the plethora of unique content sources that each firm uses today.

FINRA’s June AI note stated that “data is the lifeblood of any AI application. Accordingly, the quality of the underlying dataset is of paramount importance in any AI application. One of the most critical steps in building an AI application is to build the underlying dataset, as AI applications are best positioned to yield meaningful results when the underlying datasets are substantially large, valid, and current.”

We see this moderate view — where AI is applied as an extension of a clean, policy-controlled dataset — as ultimately prevailing for most firms.