UK financial firms are facing challenges in deploying machine learning technology, primarily due to a lack of regulatory clarity and outdated legacy systems, according to a survey conducted by the Bank of England.
Nearly half of the 71 responding firms indicated that regulations from the Prudential Regulation Authority and/or the Financial Conduct Authority hinder their ability to implement machine learning. Additionally, 25% of the participants cited a “lack of clarity” within existing regulations as a contributing factor.
Despite these regulatory limitations, the most significant obstacle to the adoption and deployment of machine learning identified by respondents was their legacy systems. Nevertheless, approximately three-quarters of firms reported either utilizing or developing machine learning applications, which are becoming more prevalent across various business areas.
This trend is expected to persist, with firms anticipating a 3.5 times increase in the median number of machine learning applications over the next three years. The insurance sector is projected to see the largest increase, followed by the banking sector.
The respondents highlighted several benefits of machine learning, including enhanced data and analytics capabilities, increased operational efficiency, and improved detection of fraud and money laundering.
While respondents generally do not view current machine learning applications as high risk, they identified some concerns. For consumers, the primary risks relate to data bias and representativeness. For firms, the main risks are associated with the lack of explainability and interpretability of machine learning outputs.