The Window logo.
How Can Financial Services Firms Evaluate the Risks of AI

How Can Financial Services Firms Evaluate the Risks of AI?

This article is originally from Fintechdrift.com, acquired by Window in December 2021.

Financial services executives are increasingly bullish on artificial intelligence.

There is an increasing tendency to stretch the definition of AI to encompass all categories of data science — so that firms can appear innovative.

It is not just PR, however. Financial firms recognize that AI can increase productivity, improve data governance, and transform how firms interact with customers.

The scope of what AI can do goes beyond improving existing processes to the development of new products that would be impossible to envision before.

But introducing AI into financial services introduces problems as well — particularly for those organizational functions concerned with controlling risk.

Problems

A major challenge is most risk managers lack an understanding of the core technology, limiting their ability to identify potential risks.

Moreover, risk managers are often not brought into the development process until after the algorithms are developed — too late for them to shape development to ensure appropriate safeguards.

This is particularly critical because “bad” AI scales quickly. An errant employee is limited by time and access, but a bad algorithm can rapidly affect many nodes of the organization.

Recommendations

With this in mind, here are three things financial services firms can do to manage risks associated with AI.

  1. Conduct an “AI Audit.” Before making any decisions, take the time to understand where AI exists in your organization, what it contributes, and its importance to the overall business. It is also worth investigating who manages the AI and where it was developed — was it built internally, or by a third party?

  2. Focus on Data First. AI is only as good as the data it uses to make its decisions. Making data reliable and secure is essential anyway, and even more so when machines make decisions based on that data.

  3. Ensure risk managers understand AI’s risks and opportunities. This doesn’t just mean that risk managers need to understand AI at a technical level. They need to understand its business applications and develop the capability to keep up with the latest technologies continually. 

Fintech Drift Team

Fintech Drift Team

Acquired by Window in December 2021, Fintech Drift was a media property that helped professionals stay ahead of tech innovations shaping the global financial markets.

More From Window

Share This Post