Ethical and design framework for Deaf-centric voice cloning

Develop an ethical and design framework for voice cloning in Deaf-centric voice technologies that addresses consent, transparency, accountability, privacy, misrepresentation, and the preservation of Deaf identity.

Background

The paper identifies voice cloning as a particularly sensitive area for Deaf-centric technologies because synthetic voices may affect how DHH people are represented and perceived. Participants raised risks involving non-consensual use, misrepresentation, privacy, stereotypes, and the preservation or removal of Deaf-accented characteristics. Although the paper surfaces user-grounded requirements such as explicit consent and transparency, it characterizes the broader ethical and design challenge as unresolved.

References

It identifies voice cloning ethics as an open research and design problem in the DHH space, establishing a set of user-grounded requirements that complement theoretical FATE frameworks such as those developed for sign language AI .

— Seeing the Voice, Preserving the Self: A Participatory Design Approach to Deaf-Centric Text-to-Speech  (2609.10199 - Atemnkeng et al., 9 Sep 2026) in Section 6, “Overall Implications and contributions,” item 5