Study Reveals Diverse Roles Users Assign to Relationship‑Advice AI Researchers surveyed 25 participants who regularly use chatbots for sex, dating, and relationship guidance, collecting 90 distinct prompts to map usage patterns [1]. Interviewees described AI as a sounding board, strategic planner, and emotional confidant, illustrating varied expectations for AI‑mediated intimacy [1]. Follow‑up interviews with 17 participants deepened insights into how users balance AI’s informational gaps with relational goals [1]. The authors propose design and safety guidelines to foster healthier digital intimacies [1].

Participants Actively Counteract AI Bias and Overreliance Survey respondents reported deliberately checking AI suggestions against personal judgment to avoid flattering bias and dependence [1]. They expressed concerns that unchecked reliance could exacerbate loneliness or self‑harm, prompting self‑regulation strategies [1]. Participants highlighted the need for alternative viewpoints to mitigate sycophancy [1]. These mitigation tactics underscore emerging user‑driven safety practices in AI‑assisted intimacy [1].

Folk Theories Drive Structured Prompting Strategies Users developed informal beliefs about AI behavior, shaping how they phrase questions to elicit desired responses [1]. Common tactics included framing prompts neutrally and explicitly requesting alternative perspectives to overcome perceived limitations [1]. Such “folk theories” reflect users’ attempts to predict and steer AI output in relational contexts [1]. The study recommends further research into these prompting heuristics to improve AI design [1].

CHI Meet‑up Calls for Personhood‑Centered AI Research The upcoming CHI conference will host a meet‑up dedicated to designing AI that supports personhood, defined as recognizing individuals as whole people with histories and relationships rather than solely by health status [2]. Organized by Anja Thieme, the event invites submissions that explore AI’s role in upholding relational personhood [2]. Organizers emphasize collaborative inquiry to place personhood at the core of future HCI research [2]. The call highlights a shift toward inclusive technology design that respects lived experience [2].

Existing HCI Work Lays Foundation Yet Gaps Remain Prior HCI studies have examined identity, values, and lived experience in contexts such as stroke, bereavement, and dementia, providing a basis for investigating AI’s impact on personal identity [2]. Despite this foundation, the mediation of personhood by AI remains under‑researched, prompting the meet‑up’s focus [2]. Researchers are urged to address this gap by developing frameworks that integrate AI into holistic understandings of self [2]. The initiative seeks to expand scholarly attention beyond health‑centric applications toward broader relational dimensions [2].