Closing the AI accountability gap

Abstract

Rising concern for the societal implications of artificial intelligence systems has inspired a wave of academic and journalistic literature in which deployed systems are audited for harm by investigators from outside the organizations deploying the algorithms. However, it remains challenging for practitioners to identify the harmful repercussions of their own systems prior to deployment, and, once deployed, emergent issues can become difficult or impossible to trace back to their source. In this paper, we introduce a framework for algorithmic auditing that supports artificial intelligence system development end-to-end, to be applied throughout the internal organization development lifecycle. Each stage of the audit yields a set of documents that together form an overall audit report, drawing on an organization's values or principles to assess the fit of decisions made throughout the process. The proposed auditing framework is intended to contribute to closing the accountability gap in the development and deployment of large-scale artificial intelligence systems by embedding a robust process to ensure audit integrity. CCS CONCEPTS • Social and professional topics → System management; Technology audits; • Software and its engineering → Software development process management.

What do we want from Explainable Artificial Intelligence (XAI)? – A stakeholder perspective on XAI and a conceptual model guiding interdisciplinary XAI research

“…Accountability is about being able to identify who is blamable or culpable for a mistake. With increasing use of artificial systems, accountability gaps might emerge [134,135]. For instance, when the use of an artificial system harms a person, it may not be clear who is accountable, as there are many parties that may have contributed to the harm.…”

Bridging the Gap Between Ethics and Practice

“…Skeptics will see these statements as self-serving corporate whitewashing, designed to generate positive public responses. However, they can produce important internal efforts such as Google's internal review and algorithmic auditing framework [92] (see Section 3.4). Corporate statements can help raise public expectations, but the diligence of internal commitments should not be a reason to limit external independent oversight.…”

Ethical principles in machine learning and artificial intelligence: cases from the field and possible ways forward

“…All these arguments suggest limitations to full disclosure of algorithms, be it that the normative implications behind these objections should be carefully scrutinised. Raji et al (2020) suggest that a process of algorithmic auditing within the software-development company could help in tackling some of the ethical issues raised. Larger interpretability could be in principle achieved by using simpler algorithms, although this may come at the expenses of accuracy.…”