The Bank of England and the Financial Conduct Authority have concluded a comprehensive year-long investigation into the utilization of data, model risk, and governance regarding artificial intelligence within the financial services sector.
The initiative, known as the Artificial Intelligence Public-Private Forum (AIPPF), was established in October 2020. It convenes a diverse group of experts from finance, technology, and academia, along with representatives from various UK regulatory bodies and governmental agencies.
Through a series of quarterly meetings and workshops, the forum aimed to facilitate dialogue, allowing regulators to grasp the practical complexities of implementing AI in financial contexts. This engagement also served to collect insights on potential principles, guidelines, and exemplary practices.
The forum underscored the significance of data integrity, model risk, and governance. It noted that issues originating at the data level can persist as model inputs, and even the most accurate data can lead to complications during the modeling phase. Furthermore, intricate AI systems present unique governance challenges.
The findings emphasize that clear regulatory expectations are vital for encouraging innovation in AI adoption. Regulators are encouraged to enhance clarity around existing regulations while ensuring that any new guidance remains flexible and includes practical case studies.
Additionally, it is recommended that regulators identify key AI use cases in the financial sector, particularly those that pose significant risks, with a view to developing effective mitigation strategies and policy approaches.
The report also advocates for ongoing collaboration between the public and private sectors, proposing the establishment of an industry consortium to tackle challenges and set standards.
For further insights, you can access the full report here.