Informing Future AI Policy Design: A National AI Strategy Policy Coherence Analysis

Abstract

As artificial intelligence (AI) continues to be globally adopted despite misuse cases like the Toronto Police Service’s Clearview AI facial recognition scandal; Cambridge Analytica’s democratic infringement, and countless criminal cases involving AI – understanding how governments’ respond to this global AI adoption remains an understudied topic (Brayne, 2020; CBC News, 2020; Dafoe, 2018; Hern, 2018). This literature gap is important to address as governments have begun responding to AI’s growing use through a range of policy initiatives.

One government response thus far has been to release national strategies that set out how a country thinks about AI as an economic, regulatory, social, or ethical issue that requires some form of public governance. One can think of these strategies as an initial indication of how countries are thinking about AI, including whether a country views AI technologies as an opportunity and/or a threat, and in what ways. However, these national AI strategies have yet to be compiled and distilled into their various policy design elements and temporally analyzed to determine their policy coherence – when various policies share several consistent and complementary policy design elements (May et al., 2006). Understanding this question of coherence is the central task of this paper.

Keywords: Artificial intelligence, policy design, policy coherence, governance, ethics