Consider this hypothetical scenario: a national car seller implements an algorithmic decision-making system to raise profits by ten percent in the next year. The algorithm will target the most competitive geographic market for a particular vehicle, recommend a price, and screen consumers for creditworthiness. The algorithm is immensely successful, and the corporation achieves its goal for increased profits. But a disturbing trend emerges when multiple consumers file lawsuits alleging discriminatory practices. Consumers claim that the corporation gave Black borrowers with otherwise equal credentials less favorable loan terms than white borrowers. Barred from bringing suit by a contractual arbitration provision and class action waiver, the consumers take to the media. In turn, shareholders demand transparency and ultimately file suit alleging securities fraud under the Securities Exchange Act.
This fact pattern may be a familiar one. It is analogous to algorithmic discrimination in fintech, employee hiring, and consumer data privacy. Consumers themselves face hurdles and contractual bars to bringing suit, and the enforcement agencies struggle to keep pace with advances in technology. Securities regulations present a final check on the untethered use of algorithmic decision-making. With modest development, securities laws can provide transparency and enforcement to deter inappropriate use of algorithmic decision systems in the corporate context. This Comment proposes three concepts to enable that development: modernize Securities and Exchange Commission (SEC) disclosure requirements, adapt the scienter and causation elements in securities fraud claims, and create a United States Patent and Trademark Office (USPTO) and SEC working group.
After scoping the problem in this introduction, this Comment proceeds in three parts. Part II evaluates corporate reliance on algorithmic decision-making. Parts III and IV take a critical look at how corporate filing requirements under the Exchange Act have adapted, or not adapted, to algorithmic decision-making. Part V proposes a new approach to advance the policy goals of the Exchange Act. I conclude with a proposal that corporations should disclose these risks in SEC filings now rather than wait for legislation. More realistically, Congress and the SEC should consider updating the SEC’s regulations.
Kevin X. Kuhn,
Algorithmic Decision-Making and Corporate Risk: Toward Transparency Through Corporate Disclosures,
100 Neb. L. Rev.
Available at: https://digitalcommons.unl.edu/nlr/vol100/iss4/7