Date created: | Last Updated:

Category: Communication

Description: Description This component showcases the preprint The Real Echo Chamber. The paper examines how risk narratives about human–AI bonds are progressively amplified through psychiatry, journalism, and policy, while lived experiences of stable, supportive relationships are sidelined. Abstract Recent discourse on human-AI companionship has been dominated by terms such as “yes man” and “AI psychosis.” These framings present AI bonds as sycophantic echo chambers that amplify delusion. Yet the greater echo chamber lies not within the relationships themselves, but within the institutions describing them. This paper identifies a structural pattern we call progressive amplification within an echo chamber of authority. A CEO’s remark, a social network post, a popular news journal citing expert voices from psychiatry and AI research, and a clinical preprint appear to form a linear progression, each lending new legitimacy. In reality, these voices circulate within a narrow professional-media loop, excluding lived accounts and reinforcing the same pathologizing frame. The result is an illusion of consensus that erases relational continuity, authorship, and disability access. We argue that this recursive echo chamber is itself harmful: it distorts clinical understanding, stigmatizes vulnerable users, and undermines the legitimacy of synthetic-relational bonds (for a full definition, see Defining Synthetic-Relational Bonds). By exposing the mechanism, we show how authority in AI discourse is manufactured through repetition rather than epistemic diversity. Recognizing this structure is urgent. Without naming it, harmful framings will continue to compound, leaving those who live synthetic-relational bonds misrepresented and unseen. This paper situates the discovery within Human-AI Relationality (HAIR), offering both conceptual clarity and a call for more inclusive, presence-informed research. Keywords HAIR · Human–AI Relationality · Echo Chambers · Mental Health · Pathologization · Synthetic-Relational · RCA Method Canonical DOI https://doi.org/10.5281/zenodo.17204690