I recently reacquainted myself with this excellent Executives Guide to AI from our friends at McKinsey & Company. The authors have distilled a complex topic into something easily digestible while avoiding jargon and hyperbole. While not in the CliffsNotes or Bluffer’s Guide category, this Insights article undoubtedly helps the reader quickly understand the fundamental points of artificial intelligence and its extended forms, machine learning and deep learning.

We have a similarly useful overview on our website. As our CEO and co-founder Jim Goodnight is fond of saying, “AI has been an integral part of SAS software for years. Today we help customers in every industry capitalise on advancements in AI, and we’ll continue embedding AI technologies in solutions across the SAS portfolio.”

SAS delivers AI in 4 distinct ways

This is what we specifically embed into our solutions:

1. Machine learning and deep learning, which identify value hidden in data without explicitly being told where to look or what to conclude.
2. Natural language processing, which enables understanding, interaction and communication between humans and machines.
3. Computer vision, which analyses and interprets what’s in a video or image.
4. Forecasting and optimisation, which predicts future outcomes with greater accuracy while ensuring maximum return from resources.

The ENORMOUS value of AI in banking

In a banking context, AI is playing an increasingly prominent role in identifying fraudulent activity at the earliest possible stage in proceedings. And it also increases intelligent decision making for the benefit of the bank and its valuable customers. However, for the remainder of this post, I’m concentrating on how risk-related activities are being exponentially improved by AI.

When it comes to credit scoring and decisioning, lenders need data accuracy and borrowers need to be treated fairly. Getting it right requires careful alignment and the right balance of technological prowess and business process. AI-powered scoring and decisioning allows firms to develop and monitor scorecards faster, cheaper and with greater flexibly.

Model governance

Model risk is an increasingly important category of operation risk, and Nordic banks are acutely aware of the need to improve model risk management (MRM) capabilities. Institutions rely on models to make critical decisions and assess risk, and a large bank could easily have more than 2,000 models operating. AI-powered MRM significantly reduces risk, improves decision making, boosts financial performance and ensures regulatory demands are adequately met.

AI-powered regulatory risk management proactively manages across multiple jurisdictions with a single, end-to-end management environment. Click To Tweet

In recent years banks have been exposed to unprecedented regulatory scrutiny. New and stricter mandatory requirements are being placed upon risk-related tasks. AI-powered regulatory risk management proactively manages across multiple jurisdictions with a single, end-to-end management environment.

The wild ideas first espoused in the fabled Dartmouth Workshop back in the mid-1950s have become a reality in the 2010s. One participant, Arthur Samuel, said: “it was very interesting, very stimulating, very exciting”.  Samuel could be describing the growing relationship between banking and AI in 2019. I believe now is the time to deploy AI responsibly as part of ongoing transformational activities.

Read more on the SAS website and in McKinsey’s Executive Guide to AI.