Beyond ethics washing: evaluating the responsiveness of US, EU, and UN AI governance to expert warnings

Abstract This study evaluates the responsiveness of the emerging global AI governance regime to the core ethical concerns articulated by the AI expert community, explicitly addressing the fundamental question of why aligning policy with expert warnings matters. Identifying the gaps between expert warnings and actual policies is critical for fostering participatory oversight and accelerating global regulatory harmonization. Utilizing a qualitative content analysis, the research examines how ethical concerns raised in the Pause Giant AI Experiments open letter are reflected or refracted across four major frameworks: the US (Donald Trump Administration) Executive Order No. 13859 [1], the US (Joe Biden Administration) Executive Order No. 14110 [2], the EU AI Act [3], and the UN’s Governing AI for Humanity Final Report [4]. Findings indicate that while the Trump administration’s market-oriented approach ‘refracts’ many expert concerns, the Biden administration, EU, and UN frameworks strongly ‘reflect’ ethical imperatives regarding human rights, safety, and accountability. However, critical gaps remain regarding the concentration of corporate power, which is often sidelined in Western regulations but more holistically addressed by the UN. Ultimately, this evaluation highlights that bridging the divide between expert consensus and varied geopolitical implementations is a necessary prerequisite for cohesive, rights-based global AI governance.