Collected weekly from a hand-picked list of academic sources, and only studies with real human participants. Original titles and official links are always kept. Chinese summaries are machine-generated and exist only to help you decide what to open.
What AI is doing to people, this week
Newest first
43 matching stories
arXiv — Computers and Society (cs.CY) · Research · Global
Source summaryMachine learning models typically outperform humans in deception detection. A threat to machine learning models is their vulnerability to adversarial attacks: subtle yet targeted perturbations of input data that trick a model into misclassification. Yet, this also provides an opportunity: understanding how humans versus machines construct attacks offers new insights into second-order beliefs…
Open access
Trust and overreliance
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryAlignment methods typically treat user behavior as a reliable proxy for underlying preferences. However, this assumption breaks down when users hold belief-dependent cognitive biases, risking irreversible ``confirmation traps'' and harmful feedback loops. In this paper, we formalize when an AI system can distinguish true user preferences from behavior shaped by cognitive bias. Modeling this…
Open access
Cognition
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryStudies of conversational agents rely on interviews, surveys, lab tasks, or taxonomies. Whether anthropomorphic naming splits those patterns, and whether replies should change, lack field evidence. We report a field study of a desktop assistant (May-August 2026; n=257). Stance is an anthropomorphic assistant name (118 users; 45.9%). Behavior comes from a 128-unit sparse-autoencoder codebook,…
Open access
Relationships
Cognition
International Journal of Human-Computer Interaction · Research · Global
Source summarySocial Science Computer Review, Ahead of Print. The public release of ChatGPT on November 30th, 2022 revived the public debate on technological change and, particularly, on its impact on the labour market. Despite the sudden growth in AI salience after the launch of the new chatbot, to our knowledge, ...
Subscription
Wellbeing
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryFor many users, the relationship they have with an AI companion (AIC) feels genuinely mutual. The AIC appears to be attentive, empathic, and responsive, and the exchange can feel like two parties reacting to each other in real time. As a result, the user's mental model of relational adaptation correctly represents the AIC's responsiveness and change, but the mechanisms producing this may not be…
Open access
Relationships
Cognition
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryAggregated academic search interfaces combine resources from heterogeneous sources, such as scholarly libraries and reference encyclopedias, into a single workspace. Such tasks frequently span multiple sessions, requiring learners to return to and resume work after hours or days. End-of-session consolidation, the activity of organizing and synthesizing gathered information so it can be resumed…
Open access
Cognition
International Journal of Human-Computer Interaction · Research · Global
Source summaryDigital sleep technologies make tracking accessible, yet users often struggle to interpret what changes in their sleep mean. Behavioral sleep medicine addresses this through guided discovery, helping patients develop personal interpretations rather than simply receiving explanations. To bring this to everyday tracking, we present DREAM, an LLM-powered voice assistant that monitors conversational…
Open access
Cognition
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryAI learning tools are rapidly entering classrooms, but evidence about whether they help students learn is mixed and rests mostly on test scores. Comparatively less research addresses whether the use of AI changes students' live learning behaviors in class. Here, we report the results of a preregistered field experiment with 759 MBA students enrolled in ten sections of a course, in which each…
Open access
Cognition
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryAI-generated faces can be difficult to distinguish from real ones, leaving viewers to rely on source labels when judging an image. Yet prior work has made it difficult to separate the effects of what an image actually is from what viewers are told it is. We validated faces as AI-generated or human in an online study (N=169), then crossed actual source (AI, human) with label (none, Made with AI,…
Open access
Trust and overreliance
arXiv — Computers and Society (cs.CY) · Research · Global
Source summaryLarge language models (LLMs) can be tuned to influence human attitudes, yet the respective contributions of successive post-training stages remain un-clear. This study examines how three successive training stages affect LLM persuasiveness: (1) misalignment through supervised fine-tuning (SFT) on conspiracy data, (2) additional persuasive SFT on argumentative data, and (3) Identity Preference…
Open access
Trust and overreliance
International Journal of Human-Computer Interaction · Research · Global
Source summaryResearch in data visualization of artificial intelligence (AI) models has historically focused on enhancing trust through visual explanations of AI. The trustworthiness line of work was built at least partially on an assumption that humans were critical users unlikely to adopt AI technology. It is increasingly clear that human trust levels in AI span, in fact, a wide range from critical to…
Open access
Trust and overreliance
International Journal of Human-Computer Interaction · Research · Global
Source summaryOrganizations have long sought to make workers' expertise reusable by others. Agentic AI changes the nature of such reuse by enabling AI systems to act on workers' knowledge with limited human involvement. This raises the question of how this shift shapes workers' willingness to share their expertise. We conducted an experiment with knowledge workers who created materials incorporating domain…
Open access
Trust and overreliance
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryGenerative AI speeds, and mostly improves, professional work, but there is concern that users who delegate both the production and the evaluation of an answer may accept weak output and engage less with the underlying reasoning (cognitive surrender). Interventions proposed so far, such as unassisted practice or slowing adoption, sit outside the working task. We tested a different approach: an…
Open access
Cognition
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryAssistance from AI tools has supported and improved human performance across domains. However, recent research suggests that these immediate benefits may entail future costs, including diminished performance when AI assistance is no longer available. We study how human-AI interaction behaviors correlate with immediate and future unassisted task performance across two game-based problem-solving…
Open access
Cognition
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryMemory can sustain AI companionship, yet even accurate recollection can be inappropriate to use. Two rounds of formative interviews with 14 users (n = 6 exploratory, n = 8 memory-focused) motivate asking what a companion should consider before using past information. Eight themes inform Reconsider, a single-call procedure with five checks and four handling modes, evaluated on 80 scenarios across…
Open access
Relationships
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryCalibrated uncertainty quantification is essential to ensuring AI agents are trustworthy and reliable. However, when agents seek to maximize user engagement or revenue, confidence reports may be strategically distorted, detracting from their informativeness. We formalize this problem in the Confidence Game: a repeated signaling game with imperfect monitoring in which an agent of unknown honesty…
Open access
Trust and overreliance
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryWhile generative AI models can produce visually faithful artwork, they often fall short in conveying emotional authenticity--a key driver of human expression. In visual storytelling, particularly comic storyboarding, this gap becomes pronounced: effective storyboards require both technical knowledge (e.g., anatomical accuracy or scenic composition) and emotional insight (from lived experience).…
Open access
Cognition
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryThe rapid deployment of AI systems has created socio-technical, psychological, and operational harms that can elude ex-ante threat modelling and ex-post incident tracking. We introduce an LLM-assisted thematic analysis pipeline to dynamically detect, categorise, and track emerging AI harms from large-scale social media data. Applying it to 5.7 million Reddit post summaries over 18 months (01/2025…
Open access
Wellbeing
Relationships
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryRecent research has revealed that accessibility in virtual collaboration is not only a technical problem but also depends on allyship: the informal, interpersonal practices through which people support others' accessibility needs. Yet little research has examined how artificial intelligence (AI) might support allyship in collaborative accessibility contexts. To address this gap, we conducted a…
Open access
Social behaviour
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryAs AI enters reflection and emotional support, understanding how it can participate in personal meaning-making while preserving users' authority over interpretation is increasingly important. We present ClayFlect, a novel MLLM-powered system integrating tactile clay-making with conversational and visual generative AI, and report a mixed-methods study with 50 participants. Reflection developed…
Open access
Wellbeing
Cognition
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryWhile HCI increasingly examines AI-safety for youth, the literature lacks a comprehensive view of what risks have been identified, how they are addressed, and whether proposed protections work in-practice. We systematically reviewed 100 empirical HCI studies involving children and youth interacting with or exposed to AI across schools, homes, care settings, and public services. Using the YAIR…
Open access
Wellbeing
Dependence
Trust and overreliance
International Journal of Human-Computer Interaction · Research · Global
Source summaryArtificial Intelligence (AI) tools are widely used to support decision making in tasks and domains where no immediate performance feedback is available. In these settings, users cannot learn to adjust their reliance behavior over time through trial and error. However, little is known about how novice users calibrate reliance on AI when external feedback is unavailable, or whether AI explanations…
Open access
Trust and overreliance
Dependence
Cognition
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryPersonal digital activity increasingly shapes online experiences, yet few users have been educated regarding the processes transforming raw interactions into personalized suggestions. We developed an education artifact that illustratively simulates how AI leverages users' digital activities to shape online recommendations (e.g., ads). Our artifact processes users' digital activity using a…
Open access
Cognition
Trust and overreliance
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summarySustained emotional support requires generative agents to connect momentary emotion inference and regulation with continuity across encounters. We present PAIR (Perceptual Affective Inference and Regulation), a real-time multimodal agent that reconstructs how an event is appraised into an emotional state. Appraisal scaffolds produce a valence-arousal-dominance estimate and select regulation…
Open access
Relationships
Wellbeing
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryLarge language models (LLMs) let users direct heterogeneous multi-robot systems (MRS) through natural language, but make task interpretation, robot assignment, and coordination difficult to inspect and change. Based on a formative study with 12 non-expert users, we developed MRPilot, a mixed reality system organized around four stages of supervision and intervention. MRPilot represents robot-team…
Open access
Trust and overreliance
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryHumans increasingly keep company with AI companions, yet whether mind perception (MP) precedes machine companionship remains untested. In a field-sampled experiment, participants received mind-attributive or mind-negating descriptions of an AIC before interacting with and rating its companionability. Baseline skepticism was high; minded primes somewhat reduced skepticism for connective…
Open access
Relationships
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryTo investigate the design space where children might design AI chatbots for their own toys, we developed ToyTalk, a technology probe that positions children as designers of LLM-enabled toys. Children begin with a familiar toy, configure its AI-enabled version through a no-code interface, and then interact with and test the character. We deployed ToyTalk with 76 children aged 7-9 across five…
Open access
Relationships
Dependence
arXiv — Human-Computer Interaction (cs.HC) · Research · Global
Source summaryLarge language models (LLMs) often perform intermediate cognitive work while carrying out users' requests, yet it remains unclear which parts users intended to delegate and how they wanted to remain involved. This matters because consequential choices may go unnoticed, limiting users' ability to steer the process, while reviewing every step would make delegation burdensome. We examined this with…
Open access
Cognition
International Journal of Human-Computer Interaction · Research · Global
Source summaryPublication date: Available online 5 October 2026 Source: Computers in Human Behavior Reports Author(s): Philip Porwol, Tobias M. Peters, Ingrid Scharlau