Banks by 2035: massive changes are on the horizon
unnamed-1_1.jpg

Although usage of artificial intelligence in the Luxembourg financial sector is currently fairly limited and still at an early stage, the grand duchy is set for wider adoption of artificial intelligence in the near future, according to a survey carried out by BCL, the central bank, and financial supervisor CSSF.

The overall level of adoption of artificial intelligence, or AI, in the Luxembourg financial sector is currently limited, as only 30 percent of the surveyed institutions currently use AI technologies like machine learning. The top five use cases are AML/fraud detection (18%), process automation (15%), marketing/product recommendation (8%), customer insights (8%) and cyber security (8%).

An important finding, according to the report, is that respondents demonstrated a high level of awareness of trustworthiness aspects of AI, considering aspects like for instance humans being in the loop, bias detection/prevention techniques, auditability and explainability.

“Several findings indicate that the adoption of AI is still at an early stage, especially regarding the implementation of advanced governance and ethical measures specific to AI,” said the BCL-CSSF report.

The study said this “confirms that Luxembourg institutions using AI are aware of the specific risks related to this technology.”

The report said its results confirm the importance of CSSF recommendations made in a 2018 white paper and the need to wait for upcoming regulation from the European Commission on AI, as presented in the Spring of 2021.

“The recent public enthusiasm for advanced generative solutions like Chat-GPT shows that the future is already here and that it will likely bring us more and more powerful AIs, with potential new ways to consume AI such as AI as a Service,” the report concluded. “it will be interesting to see how the current picture of the use of AI in the Luxembourg financial ecosystem will evolve in the coming years.”

Related articles on Investment Officer Luxembourg:

 

Author(s)
Categories
Documents
Access
Limited
Article type
Article
FD Article
No