Humans in a Loop: Ethical Agency and Speculative Pathways in UX Practice Within the Generative AI Maelstrom

The study identifies that practitioners' stance toward genAI is characterized more by ambivalence than opposition, recognizing benefits while fearing harms to cognition, labour, and well-being. Relative to prior work on the ethical dimensions of UX practice, it focuses on practitioners' lived experience in genAI product contexts and points to ambivalence rather than opposition as the closer description of that experience. Based on semi-structured interviews with nine practitioners who had experienced values misalignment, analyzed through reflexive thematic analysis and complemented by 175-word speculative microfictions as a reflexive analytic tool.

The study reports that practitioners felt significantly ethically disempowered in their UX work, due to structural barriers, the absence of ethical discussion in the workplace, and genAI-induced moral distress. It reframes ethical difficulty from a matter of individual moral judgment toward a matter of missing structural support, adding a dimension of disempowerment less developed in prior UX ethics discussion. Themes induced from interview data, combined with the author's autoethnographic perspective as a former UX practitioner and AI ethics team member.

The study documents that practitioners exerted agency through "lean in" and "lean out" tactics, oscillating between pessimism and cautious optimism. Rather than treating agency as present or absent, it presents agency as a dynamic practice of oscillation within constrained structures. Based on thematic analysis of interviews within a qualitative, interpretive design.

Using the metaphor of loops, the study argues that holding practitioners ethically accountable for outcomes they cannot meaningfully influence is increasingly untenable, yet thinking otherwise, constructing alternative realities, and self-authorship remain available, resulting in the branching-narrative design prototype HUMAN IN A LOOP: A UX Ethics Choose-Your-Own-Adventure Book. It pairs the critical argument with a design prototype that invites readers to explore ethical frictions and co-construct potential paths forward, so that speculation does not remain only at the level of argument. A combination of theoretical argument and design prototype, grounded in the interview and autoethnographic material described above.

The study applies to UX practitioners working on genAI products in the current socioeconomic climate who have experienced values misalignment; its conclusions are meant to offer an interpretive frame for understanding ethical agency, structural barriers, and the distribution of accountability, and can serve as a starting point for further design exploration and discussion. The design prototype HUMAN IN A LOOP is aimed at readers who wish to reflect on ethical frictions and co-construct potential paths forward.

A careful reader might still watch how far interviews with nine practitioners represent the wider UX community; how the interplay between the autoethnographic perspective and the interview material is handled; what role the 175-word speculative microfictions play in forming the conclusions; and how the loops metaphor and the branching-narrative prototype might be further tested and extended in later practice. Because the loaded text is at the summary level, figures and finer analytic detail are not included, which affects how fully the specific analytic steps can be grasped.