Better, Nicer, Clearer, Fairer: A Critical Assessment of the Movement for Ethical Artificial Intelligence and Machine Learning

Abstract

This paper uses frame analysis to examine recent high-profile values statements endorsing ethical design for artificial intelligence and machine learning (AI/ML). Guided by insights from values in design and the sociology of business ethics, we uncover the grounding assumptions and terms of debate that make some conversations about ethical design possible while forestalling alternative visions. Vision statements for ethical AI/ML co-opt the language of some critics, folding them into a limited, technologically deterministic, expert-driven view of what ethical AI/ML means and how it might work.

The global landscape of AI ethics guidelines

“…Fears that AI might jeopardize jobs for human workers 7 , be misused by malevolent actors 8 , elude accountability or inadvertently disseminate bias and thereby undermine fairness 9 have been at the forefront of the recent scientific literature and media coverage. Several studies have discussed the topic of ethical AI [10][11][12][13] , notably in metaassessments [14][15][16] or in relation to systemic risks 17,18 and unintended negative consequences like algorithmic bias or discrimination [19][20][21] .…”

AI Ethics – Too Principled to Fail?

“…Significant resources have been dedicated to public-private initiatives aiming to define universal high-level values or principles to guide ethical development and deployment of AI. 1 These initiatives can help focus public debate on a common set of issues and principles, and raise awareness among the public, developers and institutions of the ethical challenges that accompany AI. 2 To date, at least 63 such 'AI Ethics' initiatives have published reports describing high-level ethical principles, tenets, values, or other abstract requirements for AI development and deployment.…”

“…3 Many envision these high-level contributions being 'translated' into midor lowlevel design requirements and technical fixes, governance frameworks, and developer codes of ethics. 1 Existing initiatives to codify AI Ethics are not without their critics. Many initiatives are at least partially sponsored by industry, which has led some commentators to suggest they exist for disingenuous virtue signalling intended merely to delay regulation and pre-emptively shape the debate around abstract problems and technical solutions.…”

“…Many initiatives are at least partially sponsored by industry, which has led some commentators to suggest they exist for disingenuous virtue signalling intended merely to delay regulation and pre-emptively shape the debate around abstract problems and technical solutions. 1,4 This view undeniably has some merit: AI Ethics initiatives have thus far have produced vague, high-level principles and value statements which promise to be action-guiding, but in practice provide few specific recommendations 5 and fail to address fundamental normative and political tensions embedded in key concepts (e.g. fairness, privacy).…”

“…Framing ethical challenges in terms of design flaws ensures they remain "fundamentally technical, shielded from democratic intervention" or regulation. 1 The need to translate high-level principles and essentially contested concepts into practical requirements, which can be difficult and time-consuming in practice, is seemingly avoided.…”