Does compliance with the global anticorruption regime require the use of artificial intelligence?
Philip M. Nichols
Abstract
Business firms constantly hear that artificial intelligence has changed the world and that they must either utilize artificial intelligence or fall behind. By extension, this would be true of regulatory compliance as well as operations. This article challenges the mantra of artificial intelligence as a ubiquitous agent of change. It does so through the lens of the global anticorruption regime, a transnational web of laws, regulations, and norms that work together to rein in corruption. As this article demonstrates, the global anticorruption regime imposes on business firms a requirement to implement effective and up‐to‐date antibribery programs. Given the prevailing conception of artificial intelligence as the newly critical tool for business, it would be easy to interpret “effective” and “up‐to‐date” as requiring the use of artificial intelligence. To determine whether in fact the global anticorruption regime does, this article undertakes two analyses. First, it carefully determines the systems requirements of the type of artificial intelligence most applicable to antibribery programs—systems that can distinguish between honest and corrupt actors and transactions—and determines the regulatory constraints on the use of artificial intelligence in that way. This article then asks specifically what tasks artificial intelligence would be asked to do as part of an antibribery program, and evaluates the capacity of artificial intelligence to perform those tasks given the already determined system requirements and constraints. These analyses yield a surprising conclusion: in some instances, the use of artificial intelligence would be helpful, but for most business firms, particularly for smaller firms or firms that have not experienced bribery, the use of artificial intelligence would not be helpful and could be harmful. Regulators and legal scholars must not think of artificial intelligence as a panacea; its potential use must be analyzed in the context of objectives and the capacities, needs, and limits of artificial intelligence.
Evidence weight
Balanced mode · F 0.40 / M 0.15 / V 0.05 / R 0.40
| F · citation impact | 0.50 × 0.4 = 0.20 |
| M · momentum | 0.50 × 0.15 = 0.07 |
| V · venue signal | 0.50 × 0.05 = 0.03 |
| R · text relevance † | 0.50 × 0.4 = 0.20 |
† Text relevance is estimated at 0.50 on the detail page — for your query’s actual relevance score, open this paper from a search result.