Abstract
Artificial intelligence is no longer solely in the realm of science fiction. Today, basic forms of machine learning algorithms are commonly used by a variety of companies. Also, advanced forms of machine learning are increasingly making their way into the consumer sphere and promise to optimize existing markets. For financial advising, machine learning algorithms promise to make advice available 24–7 and significantly reduce costs, thereby opening the market for financial advice to lower-income individuals. However, the use of machine learning algorithms also raises concerns. Among them, whether these machine learning algorithms can meet the existing fiduciary standard imposed on human financial advisers and how responsibility and liability should be partitioned when an autonomous algorithm falls short of the fiduciary standard and harms a client. After summarizing the applicable law regulating investment advisers and the current state of robo-advising, this Note evaluates whether robo-advisers can meet the fiduciary standard and proposes alternate liability schemes for dealing with increasingly sophisticated machine learning algorithms.
Citation
John Lightbourne,
Algorithms & Fiduciaries: Existing and Proposed Regulatory Approaches to Artificially Intelligent Financial Planners,
67 Duke Law Journal
651-679
(2017)
Available at: https://scholarship.law.duke.edu/dlj/vol67/iss3/4