32.355 Voices on AI and Gay Dating (www.romeo.com)

Romeo surveyed 32,355 users worldwide about AI in gay dating and released an entirely generative-AI teaser (avatars and voices) plus an AI statement. Key findings: AI is already familiar—78% have tried tools like ChatGPT—but attitudes are cautious. Nearly half (49%) want dating to remain fully human; 44% reject AI match-suggestion assistants (23% yes, 33% maybe). Strong red lines emerged: 86% oppose AI accessing or chatting from private messages, 80% reject AI censorship of sexual language, and 74% would not consider an “AI boyfriend.” Yet many users accept supportive roles—around half are at least open to AI that helps (spam filtering, suggestions) without replacing people, and 68% worry AI companions could harm human connection. The results matter technically and ethically for AI/ML developers and platform designers: privacy, consent, content-moderation heuristics, and transparency are non-negotiable for this community. Age—not nationality—drives most variance (younger users are more open), and apparent country differences largely reflect demographic skews. Methodological limits apply (in-app, self-selecting sample, multiple votes allowed), so findings aren’t fully representative but give a robust signal about community sentiment. For practitioners, the takeaway is clear: focus on opt-in, privacy-preserving, assistive features with auditable moderation, and engage users in defining moral boundaries before deploying deeper conversational or matchmaking models.