Do GenAI avatars open new responsibility gaps?

Area of Science:

• Artificial Intelligence Ethics
• Human-Computer Interaction
• Moral Philosophy

Background:

• Generative artificial intelligence (GenAI) avatars, designed to represent real individuals, introduce novel ethical challenges.
• Existing literature discusses epistemic gaps in AI avatars, but the unique responsibility issues posed by GenAI avatars remain underexplored.

Purpose of the Study:

• To introduce and define the concept of
• proxy gaps
• arising from GenAI avatars.

Main Methods:

• Conceptual analysis of moral responsibility frameworks.
• Examination of the
• proxy-control paradox
• in GenAI avatar delegation.
• Identification of conditions for assigning moral responsibility.

Main Results:

• GenAI avatars create responsibility gaps by combining epistemic gaps (lack of AI knowledge) and control gaps (lack of human control).
• The
• proxy-control paradox
• highlights how seeking control leads to delegation and loss of direct oversight.
• These gaps challenge traditional criteria for moral responsibility.

Conclusions:

• Despite inherent
• proxy gaps
• , individuals can be held morally responsible for GenAI avatar outcomes.
• Responsibility is grounded in conditions related to human understanding, LLM personalization, veto rights, and outcome control.