Explanation Hacking: The perils of algorithmic recourse
This addresses ethical issues in AI explainability for users and developers, presenting a conceptual critique rather than an incremental improvement.
The paper argues that providing actionable explanations for AI decisions, known as recourse explanations, has ethical downsides and can lead to problematic explanation hacking, advocating instead for explanations focused on understanding.
We argue that the trend toward providing users with feasible and actionable explanations of AI decisions, known as recourse explanations, comes with ethical downsides. Specifically, we argue that recourse explanations face several conceptual pitfalls and can lead to problematic explanation hacking, which undermines their ethical status. As an alternative, we advocate that explanations of AI decisions should aim at understanding.