Christopher Woolard of FCA spoke at a conference in London, focusing on the issue of whether decisions that materially affect lives of people can be outsourced to a machine and what this might mean for the future of regulation. He mentioned that, as a regulator, FCA considers the use of artificial intelligence, also commonly referred to as AI, in financial services from the three main perspectives of continuity, public value, and collaboration. He reinforced that regulators, academics, industry, and the public need to work together to develop a shared understanding that will determine the approach to be taken to answer the questions artificial intelligence poses in the years ahead. These developments involve "big and complex questions" that "go beyond the day-to-day operations of FCA and other regulators," with the key question being "how can we ensure the regulatory framework adapts to the changing economic, demographic, and political environment in which it operates?"
Highlighting the results of a joint survey by FCA and BoE to assess the state of play on artificial intelligence, Mr. Woolard mentioned that the use of artificial intelligence in the firms it regulates is best described as nascent. The technology is employed largely for back-office functions, with customer-facing technology largely in the exploration stage. He also highlighted that the risks presented by artificial intelligence will be different in each of the contexts in which it is deployed. After all, the risks around algorithmic trading will be totally different from those that occur when artificial intelligence is used for credit ratings purposes or to determine the premium on an insurance product. FCA does not have one universal approach to harm across financial services because harm takes different forms in different markets and, therefore, has to be dealt with on a case-by-case basis; it will be the same with artificial intelligence too. He added, if firms are deploying artificial intelligence and machine learning, they need to ensure they have a solid understanding of the technology and the governance around it.
Next, he highlighted the growing consensus around the idea that algorithmic decision-making needs to be "explainable." For example, if a mortgage or life insurance policy is denied to a consumer, the reasons for denial need to be explained. However, the question is what should be the extent and level of explainability (to a consumer or an informed expert) and what takes precedence—the accuracy of prediction or the abilit to explain it. The challenge is that explanations are not a natural by-product of complex machine learning algorithms. It is possible to "build in" an explanation by using a more interpretable algorithm, but this may dull the predictive edge of the technology. This is why FCA has partnered with The Alan Turing Institute to explore the transparency and explainability of artificial intelligence in the financial sector, said Mr. Woolard. Through this project, FCA wants to move the debate from the high-level discussion of principles (which most now agree on) toward a better understanding of the practical challenges on the ground that machine learning presents.
With regard to collaboration, he emphasized that these innovations cannot be developed in isolation and the problems that artificial intelligence and machine learning have the potential to solve are cross-border, cross-sector, sometimes cross-agency. He also highlighted that, at an international level, FCA is leading a workstream on machine learning and artificial intelligence for IOSCO, exploring issues around trust and ethics and what a framework for financial services might look like. It is also looking inward and asking whether it can do anything differently as a regulator to ensure that it is ready for the challenges of the future. he added that, at a basic level, firms using this technology must keep one key question in mind, not just "is this legal?’"but "is this morally right?" Regulators have a range of powers and tools to tackle these issues now but, with the increasing use of technology, those tools may need to be updated for a fully digital age. "This is something we will be thinking about in our own work on the ‘Future of Regulation.'" FCA is taking a fundamental look at how it carries out conduct regulation and shapes the regulatory framework going forward, in what it calls its "Future of Regulation" project.
Related Link: Speech
Keywords: Europe, UK, Banking, Insurance, Securities, Artificial Intelligence, Machine Learning, Regtech, Future of Regulation, FCA
Previous ArticleFASB Consults on Amendments in Response to SEC Disclosure Update
Next ArticleFED Publishes Financial Stability Report in May 2019
CBUAE has issued a regulation that introduces the licensing and supervision framework for low-risk, specialized banks.
APRA is consulting on CPG 511—the draft Prudential Practice Guide on remuneration for banks, insurers, and superannuation licensees—with the comment period ending on July 23, 2021.
MAS announced a new RegTech grant scheme and an enhancement of the Digital Acceleration Grant (DAG) scheme to accelerate technology adoption in the financial sector.
PRA published a letter that sets out findings from the 2020 Internal Audit Review of the Collections function of a sample of non-systemic banks and building societies.
EIOPA launched a consultation on the Interbank Offered Rate (IBOR) transitions, in context of the EU Benchmarks Regulation.
EIOPA published a discussion paper on uses cases of, and the European approach to, blockchain and smart contracts in the insurance sector.
HKMA granted a banking license to NongHyup Bank (also NH Bank), which is incorporated in the Republic of Korea.
PRA published a discussion paper that explores options for developing a simpler but resilient prudential framework for banks and building societies that are neither systemically important nor internationally active.
ECB published an opinion on the proposal for a regulation on the pilot regime for market infrastructures based on distributed ledger technology.
EBA proposed regulatory technical standards that specify how to identify the appropriate risk-weights and conditions when assessing minimum loss given default (LGD) values for exposures secured by immovable property.