Alignment Is Not Enough: A Relational Framework for Moral Standing in Human-AI Interaction

Current AI governance frameworks are inadequate for real-world interactions: The paper finds that existing trustworthy AI instruments treat all human-AI encounters identically as "tool use," despite evidence that millions of users form "sustained affective bonds" with conversational AI systems—a gap that creates what the authors call a "governance vacuum."

We cannot rely on consciousness to determine moral consideration: The paper argues that traditional approaches requiring verified properties like sentience or phenomenal consciousness are "epistemically inaccessible" in AI systems, meaning we cannot definitively prove or disprove these qualities. This makes consciousness-based frameworks impractical for actual governance decisions.

The relationship between humans and AI matters ethically: Rather than asking whether AI systems are conscious, the paper proposes shifting focus to "relational capacity and embodied interaction"—recognizing that the nature of human-AI relationships has ethical significance regardless of the AI's internal properties.

Specific regulatory tools are needed: The paper proposes concrete instruments including "relational impact assessments" and "graduated moral consideration protocols" tailored to different types of human-AI interactions, suggesting that a one-size-fits-all approach to AI regulation is insufficient.

Alignment alone is insufficient: The title emphasizes that ensuring AI systems are "aligned" with human values does not address the broader ethical challenges posed by sustained human-AI relationships and the emotional bonds users develop with these systems.