The recently released results of a new joint survey from SAS and the Global Association of Risk Professionals (GARP) on the use of AI in risk management makes for very interesting reading. Here are the highlights from the study, which involved more than 2,000 participants from across the global financial services industry:
- 81 percent of risk professionals in financial services are already seeing the benefits of AI.
- 52 percent cite improvements in process automation.
- 45 percent cite improvements in credit scoring.
- 43 percent cite improvements in data preparation.
- 33 percent cite improvements in model validation, calibration and selection.
The report sensibly points out that “AI in its various forms – including machine learning, natural language processing and robotic process automation – is still, in terms of business applications, early-stage." But it’s also evident that firms are exploring use cases, implementing suitable solutions and planning for a future where AI is at the core of their risk-related activities. The table below is ample evidence of this.
The status of AI in risk
AI Technology | Currently Use | Planning to Use |
Forecasting | 54% | 33% |
Optimisation | 51% | 36% |
Machine learning | 34% | 46% |
Robotic process automation | 29% | 37% |
Natural language processing | 23% | 43% |
Virtual agents | 22% | 35% |
The lure of the numerous business advantages of AI is proving to be irresistible. At the top of the benefits tree is the ability to extract faster insight from data, with 78 percent of respondents citing this as a significant plus. In a tie for second, the reduction of manual tasks and improved decision making both garner support from 77 percent of risk professionals. The positives don’t end there: 73 percent of surveyed executives report higher productivity and an equal number – 66 percent – report lower operating costs and improved overall product quality and customer experience.
Caution: Mind the gap
As is often the case, these numbers don’t tell the whole story, and as might be expected it’s not all good news. A skills gap clearly exists, as 52 percent of respondents said they were at least somewhat concerned that their firms lack the necessary talent to implement and maintain AI solutions. Furthermore, there are sizable obstacles to adoption which need to be overcome. These include elements like data availability and quality (59 percent), key stakeholders’ lack of understanding of AI (54 percent) and the interpretability of models (47 percent).
With the opacity of AI being rapidly replaced by clarity over use cases, deployment options and tangible commercial rewards, it’s clear that this next-generation technology is growing up fast. Financial firms need to study best practices, not just from within their own domain, but in other vertical sectors too. It’s also cheering to observe that AI isn’t just for the sole preserve of the world’s biggest companies with the deepest pockets. As the SAS and GARP report says, “gains in the accessibility and affordability of the technology make it feasible for financial institutions below the top tier that has historically been able to fund investments in technological innovation.” Hallelujah! AI for the many, not the few.
Companies must remember to ensure their AI-powered initiatives are ethical and possess adequate oversight. We can also expect to see and hear more about FATE: Fairness, Accountability, Transparency and Explainability.
Read more about the results here.
“gains in the accessibility and affordability of the technology make it feasible for financial institutions below the top tier that has historically been able to fund investments in technological innovation.” Click To Tweet