AI has advanced radiology, yet variability across hospitals and devices undermines reliability and trust. We present a federated learning framework that combines frequency-domain harmonization and instruction-conditioned personalization to deliver consistent and interpretable diagnostic outcomes. Using FFT-based reconstructions informed by radiomics descriptors, the system reduces equipment dependency, while CLIP-based text conditioning enables clinicians to guide reconstructions to local practices and patient needs. We evaluated the framework across four hospitals with fifteen radiologists and fifty patients, spanning polyp detection, rotator cuff tear diagnosis, pneumothorax classification, and breast cancer classification/segmentation. Results show significant gains in accuracy, calibration, and robustness under cross-site transfer, without introducing prohibitive latency. Radiologists reported improved interpretability and preserved professional agency, while patients expressed greater trust, reduced anxiety, and stronger acceptance of AI involvement. This work advances a human-centered design for medical AI, aligning federated learning with transparency, equity, and trustworthy deployment.
Virtual Reality (VR) offers portable and flexible workspaces. However, enabling efficient and comfortable interactions without external input devices remains challenging. We propose leveraging redirected input to enable comfortable and touch-like interaction for quick and intuitive control. Our design study revealed that while touch interaction performs well with direct input, its performance degrades significantly under input redirection. In contrast, using pinch improves redirected input by providing self-haptic feedback and reducing input dimensionality, thereby compensating for spatial discrepancies. Based on these findings, we introduce Redirected Pinch, a bare-hand interaction technique that combines input redirection with pinch confirmation. It creates a virtual plane at waist height, remapping hand movements on the plane to a vertical window, with pinch gestures used for confirmation. A user study demonstrated that Redirected Pinch achieves a strong balance of accuracy, efficiency, comfort, and sense of agency across fundamental interactions.
Parametric modeling is a prevailing 3D modeling approach in design, architecture, and engineering. The emergence of multimodal large language models (LLMs) brings a new opportunity to lower the entry barriers to this powerful tool. However, describing 3D geometries through natural language can be fuzzy and challenging. We introduce co-speech gesture, a natural and expressive interaction modality to complement text prompts for LLM-empowered generative parametric modeling. We first conducted an elicitation study to explore and categorize co-speech gesture expressions. Based on the findings, we designed a multimodal fusion pipeline that parametrizes gestures and synthesizes them with speech. This approach reduces language ambiguity by translating implicit user intentions into explicit parametric attributes, thus lifting the model generation performance. We conducted a two-session user study testing and comparing it with traditional language and sketch inputs. This work streamlines the parametric modeling workflow and explores novel multimodal interaction paradigms for LLM-empowered design and creation. Project page: https://rl-duan.github.io/JustShape/
Gaming is a meaningful part of children’s lives, yet its safety has drawn increasing concerns from scholars and the public. On platforms like Roblox, children may encounter extremist roleplay, scams, or virtual rape. Prior research has emphasized technical interventions that address risks after they occur and ethical frameworks for game design, but children’s perspectives on safety design remain missing. To address this gap, we conducted a cooperative inquiry study to co-design safety mechanisms with 22 children aged 7–12. Children proposed designs emphasizing transparent information about games and purchases, community accountability through reporting and reviews, player empowerment to manage social boundaries and engagement, and age-appropriate game navigation. Our findings extend safety-by-design research by foregrounding children’s perspectives, showing how they envision safety mechanisms across both game and platform design, while enjoying safe play through risk exposure, allocating trust, and balancing platform support with agency.
Teen participation in esports is rapidly expanding, raising concerns about how competitive gaming shapes adolescent mental well-being. Existing mental well-being initiatives often adopt adult-centered approaches that overlook teenagers’ lived realities. This study explores mental well-being from the perspective of teen esports players. Through nine participatory design sessions with 34 participants, including adolescent players, coaches, and program coordinators, we examined how teens conceptualize a “healthy player” and sustain mental well-being in gaming. Findings emphasize three key insights: (1) for teens, being “healthy” means winning together, where well-being is tied to collective outcomes and social responsibility; (2) most stressors stem from esports environments beyond their control, underscoring the need for emotional resilience; and (3) teens favor simple, everyday coping strategies, such as taking breaks, reframing losses, adjusting play environments, and drawing on peer encouragement, over formal programs. These patterns resonate with Cognitive Behavioral Theory, suggesting that cycles of thought, emotion, and behavior underpin resilience. We argue for youth-centered, culturally relevant mental well-being strategies and micro-interventions embedded in the daily practices of adolescent esports.
Post-Roe, people capable of pregnancy face fragmented reproductive privacy landscapes in the United States (U.S.), with risks spanning legal, digital, and interpersonal domains. These conditions demand new forms of privacy guidance. We analyzed 212 reproductive health zines— a DIY, subversive, and collectively produced media genre—to understand how they communicate reproductive health information. Zines foreground embodied, first-person narratives interwoven with historical context, medical guidance, and activist messaging. We argue their use of subversive or alternative medical knowledge enhanced credibility in contexts of low institutional trust. While some zines offer digital privacy strategies, many focus on avoiding institutional exposure altogether. These emotionally resonant, context-sensitive accounts illustrate threat models attuned to entangled risks of interpersonal betrayal, legal precarity, and surveillance. We conclude with design implications for how zines might better support people navigating reproductive risk through what we call narrative threat modeling—a situated practice that communicates privacy strategies through story, tone, and form rather than technical instructions or prescriptive checklists.
While the rise of AI has benefited professionals, it also induces technostress that threatens their expertise and jobs. To ensure the human-centered advancement of technology, a deep understanding of users technostress and how to cope with it is essential. Despite technostress having long been discussed, the growing integration of AI tools into professionals’ everyday work amplifies these challenges and calls for further exploration. Accordingly, this is a timely moment to examine their real-world experiences and voices. Thus, our study aims to investigate AI-induced technostress experienced by professionals, and the coping strategies they employ. Through focus group interviews with 19 professionals from diverse fields, we identified seven AI-induced technostressors and examined their coping strategies along two dimensions: stress Coping Style (problem-focused and emotion-focused) and Value Orientation (AI-oriented and humanness-oriented). Drawing on professionals’ coping strategies, we suggest practical implications to support users in coping with AI-induced technostress.
Large language models (LLMs) are increasingly deployed, yet they introduce significant privacy risks by disclosing personally identifiable information (PII) during interactions. Although prior work has demonstrated the feasibility of extracting PII from LLMs, no comprehensive study has evaluated the actual extent of PII leakage across mainstream LLMs or investigated user perceptions, literacy, and behavioral responses to these risks. To address these gaps, we conduct a large-scale evaluation of PII leakage in popular LLMs, demonstrating that attackers can extract email addresses and phone numbers with high success rates. Through a mixed-methods study involving 20 interviews and 204 survey participants, we identify significant discrepancies between user concerns and behavior: despite strong concerns about PII leakage and limited understanding of training data provenance, users continue to use LLMs due to perceived utility, often exhibiting privacy cynicism. Based on these findings, we propose design implications for enhancing the privacy-utility balance in future LLM deployments.
With generative AI enabling easier production of sexually abusive content, deepfake sexual abuse has intensified, making anyone with visual data be a potential victim or perpetrator. Current moderation systems for non-consensual intimate imagery (NCII) are platform-centric, reactive, and poorly aligned with the workflows of real-time monitors and survivor supporters. To address this gap, we held participatory design workshops with 10 activists affiliated with victim advocacy and survivors experienced in combating deepfake sexual abuse in South Korea. Their insights revealed distinctive challenges, including ambiguity in content classification, barriers to evidence collection, and increased workloads and safety risks during monitoring. Participants suggested features for proactive protection, long-term case tracking, and cross-platform coordination, while emphasizing the need for conversations about data ownership and platform accountability. Based on these findings, we discuss design implications for system and policy that foster multi-stakeholder collaboration to prevent harm, strengthen cross-platform response, and reduce secondary trauma for activists.
An exponential rise in manuscript submission volume has strained the peer review system, prompting interest in automation from overburdened scholars and publishers. We systematically evaluated GPT-4.1, GPT-4o, o1, o3, o3-mini, and GPT-5 as “peer” reviewers, comparing their evaluation and acceptance of 137 manuscripts from an open dataset (PeerRead) with the corresponding human-generated reviews. While o3’s and GPT-5’s acceptance rates were close to the human benchmark (∼ 67% of submissions), others approved nearly every paper (>98%); all models performed extremely poorly on accuracy, precision, and recall metrics. To probe this striking “yes-bias”, we profiled the LLMs using Schwartz’s Portrait Values Questionnaire (PVQ-RR) and found that all LLMs emphasized self-transcendence and openness-to-change and de-emphasized conservation and self-enhancement. We argue that value orientations of LLMs we investigated are misaligned with the values underpinning peer review, and suggest new research on aligning AI judgment systems with human goals in this context.
Data powering AI is often opaque. Researchers, NGOs, and law and policy leaders have called for greater transparency about how data is used for training, fine-tuning, and evaluation. While data transparency is often championed as crucial, what it concretely enables is largely implicit. Similarly, the concerns developers seem to have about transparency go unstated. This lack of clarity has led some researchers to critique transparency demands as disconnected from the actual benefits—or risks—to specific stakeholders. We analyze documentation from four stakeholder groups to create a taxonomy of the risks and benefits of dataset transparency. Data transparency is perceived as either a risk or a benefit given a stakeholder’s position, rather than wholesale. We also propose data availability and data documentation as two lenses through which to consider transparency. We discuss how best to strategically promote situational data transparency that takes into account the relationship between stakeholder position, transparency modality, and benefits/risks.
Perceptual grouping enables people to organize elements into units according to intrinsic (e.g., proximity) and extrinsic (e.g., common region) principles. However, the role of physical surfaces as extrinsic grouping cues for virtual elements in Augmented Reality (AR) remains unclear. To provide a deeper understanding, we conducted two within-subject studies. The first study (N = 24) using repetition discrimination tasks revealed that surfaces can be common-region cues in 3D, with their influence depending on their distance to target objects along the viewing direction. Building on these findings, the second study (N = 24) employed both objective and subjective measures to capture the interaction between proximity and common-region cues in AR. Results indicate that competing cues reduce group clarity. They also enable us to distill people’s strategies for improving the clarity by leveraging their physical and virtual environments. Finally, we propose design recommendations for future AR systems in assisted grouping tasks.
People’s online information processing is strongly shaped by social influence, and large language models (LLMs) now enable social bots to manipulate such influence at scale. This paper examines the effects of LLM-driven adversarial social influence—a strategy in which automated agents employ LLMs to distort truth by making misinformation appear credible or by undermining factual news—on how people evaluate and share information. Across two pre-registered, randomized experiments, we first show that exposure to LLM-driven adversarial social influence significantly reduces people’s ability to judge the veracity of news and lowers their discernment between sharing true versus false content. We then test two credibility prompts: AI-generated content detectors and warnings, as potential interventions. Results show that both prompts mitigate some harms such as by improving misinformation detection, though their effectiveness were dependent on the context. We conclude by discussing the risks of LLM-driven adversarial social bots and the implications for designing interventions to combat misinformation.
Beyond hallucinations, Large Language Models (LLMs) can craft deceptive arguments that erode users’ critical thinking, posing a significant yet underexamined societal risk. To address this gap, we develop a taxonomy of eight deceptive persuasion strategies by integrating top-down rhetorical theory with a bottom-up analysis of 3,360 AI-generated messages by four LLM families and examining their effects on user perceptions. Through a large-scale user study (N=602) complemented by a think-aloud protocol, we found that participants were vulnerable to Information Manipulation and Uncertainty Exploitation, especially when a message contradicted their prior beliefs. Vulnerability was significantly higher for participants with low cognitive reflection, low topic knowledge, and low topic involvement. Qualitative analyses further revealed that participants were persuaded by the plausibility of an overall narrative even when they distrusted specific details, interpreting deceptive outputs as logically framed information that broadens perspective. We discuss critical implications of these findings for the design of trustworthy AI systems, adaptive user interfaces, and targeted literacy education.
As generative AI (GenAI) is integrated into everyday technologies, it offers new accessibility opportunities and risks for disabled people. However, little is known about how disabled people navigate GenAI in their everyday lives, particularly how trust, privacy, and intersectional identities affect these experiences. We present findings from seven cross-disability focus groups (N=20) that explore how disabled people navigate GenAI. Our findings reveal that while GenAI supports autonomy, efficiency, and communication, it also introduces accessibility taxes and ethical dilemmas. Although participants voiced skepticism, many continued using GenAI out of necessity. Finally, we found identity-based benefits and tensions, in which GenAI preserved and validated intersecting identities, but also misrepresented and erased those identities. We frame these negotiations as a constant balancing act between access and risk, urging research to further examine how “access” is conceptualized. We offer implications for creating GenAI tools that are transparent, trustworthy, and responsive to intersectional identities.
In designing for dementia, the concept of person-centered care (PCC) effectively shifts attention away from deficit orientations. However, it predominantly focuses on the roles of humans in sustaining personhood, leaving the roles of nonhuman actors in care assemblages under-explored. To address this, we propose a theory–method package that extends PCC with posthumanist perspectives. We applied it in a six-month participatory design process to develop a conversational agent (CA) with four people living with dementia (PlwD), their relatives, and care workers. We report both the design process and analysis of conversations between the CA and one PlwD. Tracing the socio-material and ethical assemblages that emerged in these intra-actions, we identify moments of recognition, validation, holding, and facilitation. Situating these within broader discussions of care, process-oriented ethics, and posthuman design, we illustrate opportunities and limits to design for meaningful experiences of personhood between people with dementia, CAs, and other nonhuman actors.
As AI-assisted coding becomes standard in software development, computer science educators need a clearer understanding of how Large Language Models (LLMs) can support the learning process. Recent work has examined how students can benefit from using LLMs in their courses, but most studies rely on self-reported usage or controlled experiments with short, isolated programming tasks. To complement these approaches, this paper investigates how students organically leverage LLMs in an advanced computer science course where assignments reflect real-world complexity. We analyze 448 LLM chat logs from 147 students across two offerings of a senior-level web programming course at a large U.S. research university. Through open coding, we identified 14 distinct prompt–response pair types that cluster into three categories: to generate code, debug code, and explain programming concepts. Our analysis reveals that how students interact with LLMs correlates with academic performance. High-effort detailed specifications for code generation positively correlated with final grades (r = 0.25, p < 0.01), whereas low-effort behaviors such as pasting raw error messages showed negative correlations (r = −0.34, p < 0.01). We also observed a temporal shift toward explanation-oriented interactions, suggesting that students increasingly use LLMs as conceptual tutors and not just as code generators.
While memes enhance social interaction on social media, they can raise privacy and security concerns. Despite research on overtly toxic or unsafe memes, little attention has been given to users’ experiences with seemingly safe memes and how contextual factors trigger privacy concerns. This study explores users’ comfort levels, influencing factors, underlying reasons for discomfort, and unmet needs when engaging with such memes. We first collected and analyzed 2,317 Reddit posts describing real-world meme experiences, then conducted an online survey with 324 participants to evaluate comfort across curated scenarios. Our findings reveal that perceived-safe memes can cause harm when shared inappropriately, with comfort shaped by content and context. Privacy concerns intensify with deeper involvement, strangers, and sensitive meme topics. We identified users’ desire for consent and control in meme interactions. Based on our study, we make recommendations for users, developers of social media platforms and policymakers to address meme-related privacy and contextual concerns.
The social media app BeReal positions itself as a space for meaningful connections, however, little is known about how the app's unique combination of ephemerality, informality, and improvisation actually supports relationship maintenance. We aimed to understand what role BeReal plays in young adults’ friendships and lives. Drawing on interviews with 31 young adults at a large university in the northeastern U.S., we find that users treated BeReal as a fun, low-effort space to share glimpses of everyday life with smaller networks of friends. BeReal helped users maintain relationships, especially with past or geographically distant friends but not necessarily deepen bonds with close friends. Users welcomed the app's minimalistic user experience but raised doubts about the platform's longevity. Based on our findings, we present the Social Media Effort (SME) heuristic to help designers and researchers visualize how content and audience shape the social media ecosystem. We advocate that the HCI community design new platforms, since dominant business models are not poised to support relationship maintenance.
