•  
  •  
 

AIS Transactions on Human-Computer Interaction

Abstract

Generative AI systems increasingly shape how users interpret, author, and judge. However, prevailing approaches to AI design locate ethical integrity upstream in models, datasets, guardrails, and compliance regimes, and offer limited guidance on how ethical commitments are enacted at the interface where consequential interaction unfolds. We argue that ethical integrity in generative AI is irreducibly interactional. We introduce interactional integrity as a defining property of human–AI interaction, which reflects the extent to which ethical agency is preserved, exercised, and sustained throughout its use. Drawing on Sartrean ethics, we translate four governing principles—freedom to choose, fidelity to values, fidelity to purposes, and foresight of consequences—into interface-level design requirements. We operationalize these requirements through the Interactional Integrity Framework for Generative AI (IIF-GenAI) design, which specifies four classes of interactional affordances: invocation, imprint, vector, and resonance. Together, these affordances preserve the conditions under which users can initiate, author, align, and remain answerable for AI-mediated action. Theoretically, the study bridges existential ethics and affordance theory to conceptualize interactional integrity as a constitutive quality of ethical human–AI interaction. Practically, it offers a design vocabulary for developing generative AI systems that preserve meaningful choice, support authentic expression, sustain intentional engagement, and strengthen responsibility in use.

DOI

10.17705/1thci.00248

Download

Share

COinS
 

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.