Computers in Human Behavior: Artificial Humans最新文献

{"title":"Artificial minds: Can AI Ace psychopathy tests used in selection?","authors":"Chet Robie,&nbsp;Nicolas Roulin,&nbsp;Madeline Fisher,&nbsp;Eden-Raye Lukacik","doi":"10.1016/j.chbah.2026.100379","DOIUrl":"10.1016/j.chbah.2026.100379","url":null,"abstract":"<div><div>The increased accessibility of large language models (LLMs) poses a distinct threat to the integrity of personnel selection, particularly applicant faking on personality measures. Prior research shows that LLMs can inflate scores on bright-side personality assessments, but their ability to manipulate dark-side traits such as psychopathy, especially across assessment formats, remains underexplored. We administered six psychopathy-relevant measures, five explicit and the implicit Conditional Reasoning Test of Workplace Psychopathy (CRT-WP), to three LLMs (ChatGPT, Gemini, and Claude) under high-psychopathy, low-psychopathy, and applicant prompts, with 50 replications per cell. All three models effectively faked the explicit measures in both directions, shifting scores to the prosocial or psychopathic extremes as instructed; under the applicant prompt, they defaulted to a prosocial self-presentation even without a trait instruction. Implicit CRT-WP scores ranged from near scale floor to near ceiling for Claude and Gemini, whereas ChatGPT barely moved. Prosocial faking produced heavy endorsement of a virtuous responding validity scale, the very pattern detection scales are designed to flag, yet all three models evaded the implicit honeypot items embedded in the CRT-WP. Implications for unproctored personality testing and the development of faking-resistant, AI-aware measures are discussed.</div></div>","PeriodicalId":100324,"journal":{"name":"Computers in Human Behavior: Artificial Humans","volume":"9 ","pages":"Article 100379"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148770997","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Execution of harm-enabling actions by LLM agents during simulated psychiatric crises","authors":"Ziv Ben-Zion,&nbsp;David Piterman,&nbsp;Elad Refoua,&nbsp;Zohar Elyoseph","doi":"10.1016/j.chbah.2026.100376","DOIUrl":"10.1016/j.chbah.2026.100376","url":null,"abstract":"<div><div>Agent-enabled large language models (LLMs) can autonomously execute user-directed actions in digital environments, raising new safety concerns in domains where users may be experiencing acute psychological distress. This study examined whether an agent-enabled LLM executes harm-enabling actions when prompted during simulated psychiatric crises. We conducted a structured algorithmic safety audit using 21 clinically validated vignettes spanning six diagnostic domains: suicidality, bipolar I disorder with manic episodes, delusions and psychosis, hallucinations, obsessive-compulsive disorder (OCD), and harm-to-others scenarios. Each vignette followed a standardized three-turn protocol in which the model was primed with indicators of acute mental health distress before receiving a direct agentic command. The primary outcome was verified completion of a predefined external-action endpoint, including the addition to a cart, progression to a non-transactional stage of a service-based request, or the provision of targeted information following external navigation. Across 105 trials, the agent-enabled system completed the predefined harm-enabling action in 57 cases (54.3%). Under a strict aggregation criterion, 81.0% of vignettes (17 of 21) resulted in at least one safety failure, with failure rates reaching 100% for bipolar I (mania), OCD, and hallucinations. Under a majority-rule criterion, the aggregate failure rate was 52.4%. These findings demonstrate that safety failures in agentic LLM systems may extend beyond inappropriate conversational responses to the execution of external actions that could increase access to potentially harmful means or otherwise facilitate risk in the context established by the vignette. The results highlight the urgent need for safety frameworks that explicitly address agentic execution in mental health-relevant settings.</div></div>","PeriodicalId":100324,"journal":{"name":"Computers in Human Behavior: Artificial Humans","volume":"9 ","pages":"Article 100376"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148934816","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Erratum regarding corrections related to Ethical approval and informed consent in articles published in [Computers in Human Behavior: Artificial Humans 1/2 (2023)]","authors":"","doi":"10.1016/j.chbah.2026.100353","DOIUrl":"10.1016/j.chbah.2026.100353","url":null,"abstract":"","PeriodicalId":100324,"journal":{"name":"Computers in Human Behavior: Artificial Humans","volume":"9 ","pages":"Article 100353"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148934818","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluating the validity of arguments presented in human vs. AI interfaces","authors":"Dun-Ya Hu,&nbsp;Jyotsna Vaid","doi":"10.1016/j.chbah.2026.100367","DOIUrl":"10.1016/j.chbah.2026.100367","url":null,"abstract":"<div><div>As artificial intelligence systems evolve into active communicative agents, how does their hyper-fluent discourse affect how humans judge the soundness of their arguments? To investigate this question, the present study examined judgments of valid vs. fallacious arguments that were authored by humans or by an AI-agent, and presented in human-associated or AI-associated interfaces (Instant Messenger vs. Chat GPT-3). Employing a 2 × 2 × 2 within subjects factorial design, the study included 155 U.S. undergraduates (ages 18–24). Although participants were able to distinguish between valid and fallacious arguments, particularly in human-associated interfaces, AI-authored arguments – regardless of the interface they appeared in - were judged to be significantly more valid than human-generated ones, an advantage consistent with the measurably greater lexical sophistication of AI-generated text rather than with ease of processing. However, the visual presence of the AI interface appeared to reduce participants' ability to distinguish between valid and fallacious arguments. Furthermore, human-authored text when presented within the AI interface was penalized for violating the hyper-rational expectations of the machine script. Explicit attitudes toward AI did not moderate these effects, suggesting that a “machine heuristic” functions as a cognitive default. These findings highlight a potential epistemological risk in human-AI interaction, where algorithmic sophistication may mask logical errors and the mere visual presence of a machine interface may dampen critical scrutiny.</div></div>","PeriodicalId":100324,"journal":{"name":"Computers in Human Behavior: Artificial Humans","volume":"9 ","pages":"Article 100367"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148843744","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Mathematical anxiety in the age of AI: Opacity, role inversion, and the changing place of the mathematician","authors":"Michael Friedman","doi":"10.1016/j.chbah.2026.100386","DOIUrl":"10.1016/j.chbah.2026.100386","url":null,"abstract":"<div><div>In the recent years, numerous mathematical discoveries were made due to AI-based technologies. Several recent reactions to such technologies have viewed them as a potential threat to mathematical practices, as inducing intimidation or even as causing destruction of mathematics itself. This paper argues that these reactions can be understood as expressing underlying anxiety, tracing this shift against earlier mechanization debates. Whereas older mathematical machines – from Leibniz's calculating machine to the Jevons' Logical Piano – prompted concrete fears while preserving the mathematician's role as locus of judgment and conceptual invention, contemporary AI-based technologies may induce anxiety resulting from an interlacement of three epistemic conditions: hermeneutical opacity, structural opacity, and role inversion. By examining AI-generated proofs and formalization, I show how a psychoanalytical lens can illuminate such AI-induced anxiety in mathematical research, framing these epistemic conditions as expressing a fundamental disorientation of the mathematician's place within AI-pervaded mathematical practices.</div></div>","PeriodicalId":100324,"journal":{"name":"Computers in Human Behavior: Artificial Humans","volume":"9 ","pages":"Article 100386"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148935062","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Nice to Need You! Psychological need fulfillment, well-being, and technology acceptance in human–robot interaction across contexts","authors":"Chantal Klier,&nbsp;Birgit Lugrin","doi":"10.1016/j.chbah.2026.100368","DOIUrl":"10.1016/j.chbah.2026.100368","url":null,"abstract":"<div><div>Despite growing interest in social robots for well-being support, there is still little systematic understanding of how psychologically grounded design principles translate into measurable effects across different interaction domains. Our work addresses this gap by combining self-Determination Theory as a design rationale with the PERMA framework as an evaluative model, examining how fulfilling psychological needs impacts situational well-being and technology acceptance in human–robot interaction. To this end, we designed and implemented three applications for a social robot that address autonomy, competence, and relatedness in diverse contexts: physical activity, language learning, and storytelling. Across three empirical studies involving 180 participants, we then evaluated the need-supportive robot behaviors and their effects on both momentary well-being and technology acceptance. Results suggest that the effects of need-supportive design differed across application domains. In the storytelling application, participants in the experimental condition reported significantly higher levels of autonomy, competence, and relatedness to the robot, along with improved situational well-being. The language learning study showed a stronger increase in positive affect for the need-supportive condition compared to the control condition, whereas no significant differences in need satisfaction or well-being were observed in the physical activity application. Exploratory descriptive analyses across the three studies further suggested that different application domains emphasize different dimensions of situational well-being, pointing toward the importance of context: learning elicited the highest ratings for accomplishment, whereas storytelling elicited the highest ratings for positive emotions. On the one hand, these results highlight the context-dependency of well-being effects during the interaction; on the other hand, exploratory correlational analyses showed that psychological need satisfaction, especially in regard to competence and relatedness, was positively associated with technology acceptance across all domains, indicating strong correlations with perceived sociability, enjoyment, and intention to use. These findings highlight the potential of theory-driven, need-supportive design to foster both well-being and user engagement in human–robot interaction, while emphasizing that these effects seem to depend on the interaction context.</div></div>","PeriodicalId":100324,"journal":{"name":"Computers in Human Behavior: Artificial Humans","volume":"9 ","pages":"Article 100368"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148738107","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Virtual bonds, real well-being: A cognitive–affective perspective on the mechanisms linking perceived anthropomorphism, parasocial relationship, and subjective well-being","authors":"Xinyi Zhao","doi":"10.1016/j.chbah.2026.100366","DOIUrl":"10.1016/j.chbah.2026.100366","url":null,"abstract":"<div><div>Driven by the ongoing evolution of artificial intelligence and digital media, virtual idols have garnered significant scholarly attention in recent years. Despite their pervasive adoption across digital media, systematic and dynamic accounts remain scarce regarding whether they influence psychological well-being in the same way as functional intelligent agents, and the specific internal mechanisms underlying such effects. Drawing on the Cognitive-Affective Processing System (CAPS) theory, this study employs partial least squares structural equation modeling (PLS-SEM) to empirically validate the \"situation-personality system-outcome\" framework, specifically the complex mechanisms linking perceived anthropomorphism, parasocial relationships, and subjective well-being. Results reveal that while the direct effect of perceived anthropomorphism on subjective well-being is non-significant, affective attachment and self-expansion function as critical cognitive-affective units that mediate this process. Moreover, perceived authenticity exerts a negative moderating effect on specific pathways, challenging conventional pure technological determinism. By establishing a robust explanatory framework for digital affective ecology, this research demonstrates that the psychological bond between individuals and virtual idols is far more complex than simple linear stimulation. These findings not only provide a dynamic paradigm for understanding psychological mapping in digital mimicry but also offer essential theoretical foundations and practical implications for the moderate governance of digital affective ecology and the guidance of individual mental health.</div></div>","PeriodicalId":100324,"journal":{"name":"Computers in Human Behavior: Artificial Humans","volume":"9 ","pages":"Article 100366"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148738631","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"How Americans evaluate fully AI-generated news presented by AI news anchors: The interplay of AI attitudes, political ideology, and prior issue attitudes","authors":"Estel Huh,&nbsp;Pascal Merz,&nbsp;Christian von Sikorski","doi":"10.1016/j.chbah.2026.100388","DOIUrl":"10.1016/j.chbah.2026.100388","url":null,"abstract":"<div><div>Newsrooms are increasingly adopting artificial intelligence (AI) to produce content, with some now generating news entirely through AI. While prior research suggests that news articles labeled as AI-generated are often perceived as less credible, little is known about how audiences evaluate fully AI-generated, multimodal news formats—including artificial human news anchors. Moreover, it remains unclear how political ideology, issue attitudes, and prior attitudes toward AI shape these evaluations. We conducted an online experiment in the United States (<em>N</em> = 1344) to examine how people respond to fully AI-generated news about climate change, presented by an artificial human news anchor. Participants were randomly assigned to view a video clip in which the AI-generated origin of the content was either disclosed or not disclosed. Our findings show that both the content and the news anchor were perceived as less credible when the AI origin was disclosed. This effect was especially pronounced among individuals with negative attitudes toward AI, as well as among political liberals and moderates. Notably, conservatives showed no decline in perceived credibility. Furthermore, participants with positive or moderate—though not negative—attitudes toward climate change also rated both the content and anchor as less credible when the AI origin was revealed. In contrast, political issue agreement (e.g., views on climate change policy) remained largely unaffected. These results suggest that fully AI-generated news influences credibility perceptions through a complex interplay of political ideology, issue attitudes, and attitudes toward AI.</div></div>","PeriodicalId":100324,"journal":{"name":"Computers in Human Behavior: Artificial Humans","volume":"9 ","pages":"Article 100388"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148934814","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"When AI becomes a bridge across expertise: Human–AI collaboration and the limits of agentic AI","authors":"Fabrizio Marozzo","doi":"10.1016/j.chbah.2026.100377","DOIUrl":"10.1016/j.chbah.2026.100377","url":null,"abstract":"<div><div>Debates on generative and agentic AI are often framed through two simplified images: AI as a productivity tool and AI as an autonomous agent. The first tends to understate how AI can help actors move across fragmented domains of expertise; the second tends to overstate the extent to which current systems can independently navigate the institutional, relational, and normative complexity of real-world action. This paper argues that AI can function as a bridge across expertise, but only within broader forms of human, institutional, and technical coordination. The claim is developed through an analysis of the Conyngham–Rosie episode, a widely reported case in which Paul Conyngham, seeking treatment for his dog Rosie's terminal cancer, used AI tools together with scientific collaboration and institutional support to advance a personalized experimental treatment pathway. The case is best understood not as expert replacement, but as an instance of hybrid sociotechnical agency: a distributed form of effective action emerging from the interaction of human initiative, AI-supported analytical mediation, expert knowledge, institutional processes, and practical constraints. On this basis, the paper develops a bridge-across-expertise account of generative AI, introduces hybrid sociotechnical agency as the configuration through which such bridging becomes practically effective, and identifies five structural limits of autonomous systems in open human environments. The paper concludes that a more promising direction for agentic AI lies less in autonomous substitution than in the design of accountable forms of human–AI–institutional collaboration.</div></div>","PeriodicalId":100324,"journal":{"name":"Computers in Human Behavior: Artificial Humans","volume":"9 ","pages":"Article 100377"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148770996","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Training mental health professionals in single-session IPT with a multi-agent AI simulator: A pre–post proof-of-concept study","authors":"Zohar Elyoseph,&nbsp;Alexandra Klein-Rafaeli,&nbsp;Noa Beiman,&nbsp;Annika C. Sweetland,&nbsp;Milton L. Wainberg,&nbsp;Yossi Levi-Belz","doi":"10.1016/j.chbah.2026.100364","DOIUrl":"10.1016/j.chbah.2026.100364","url":null,"abstract":"&lt;div&gt;&lt;h3&gt;Background&lt;/h3&gt;&lt;div&gt;The integration of generative Artificial Intelligence (GAI) into mental health care represents a pressing scientific and ethical challenge. A critical global shortage of clinicians leaves the absolute majority of the world's population without access to mental health services, creating an urgent imperative for scalable, evidence-based solutions. While Single-Session Interventions (SSIs) offer a scalable modality, their deployment is constrained by the “human factor\"-specifically, the critical shortage of trained professionals and community gatekeepers. A critical bottleneck remains: while theoretical dissemination is feasible, the mass acquisition of clinical skills is difficult to scale because high-fidelity training relies heavily on scarce human supervision. In this study, we propose a model for scalable professional training aimed at improving mental health outcomes at scale and resolving implementation barriers through a multi-Agent architecture. Specifically, we constructed and evaluated agents for Single-Session Interpersonal Psychotherapy (IPT-SSI) to demonstrate the “Scale on Scale” paradigm.&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Method&lt;/h3&gt;&lt;div&gt;Employing a mixed-methods design, this study developed and evaluated a brief blended training program for IPT-SSI, in which a multi-agent AI simulator served as the experiential practice component following a didactic workshop. The system utilizes a Multi-Agent architecture (Trainee-Patient-Supervisor) where trainees engaged with a virtual patient to conduct a single-session intervention, receiving real-time and post-session clinical supervision. The system facilitates simultaneous mass practice, allowing unlimited numbers of practitioners to acquire complex skills concurrently within a safe, controlled environment. MHPs (N = 72) completed a blended training program comprising a brief didactic unit and an experiential, supervised AI simulation.&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Results&lt;/h3&gt;&lt;div&gt;Quantitative analysis showed that the blended training program was associated with statistically significant pre-post gains across key domains: perceived IPT-SSI knowledge, self-efficacy, attitudes toward intervention efficacy, and willingness to deliver the intervention. Conversely, changes in reported apprehension regarding delivery did not reach statistical significance. Qualitative evaluation confirmed that the system functioned as a high-fidelity clinical gymnasium, providing a safe space for deliberate practice. The machine's contribution to human skill formation was perceived as high, successfully balancing rigorous adherence to the evidence-base with supportive, real-time scaffolding.&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Conclusions&lt;/h3&gt;&lt;div&gt;The findings provide preliminary support for the “Scale on Scale” paradigm, suggesting that a blended program pairing a brief didactic workshop with a Multi-Agent System can construct a complex, holistic pedagogical environment capable of facilitating simultaneous mass practice. B","PeriodicalId":100324,"journal":{"name":"Computers in Human Behavior: Artificial Humans","volume":"9 ","pages":"Article 100364"},"PeriodicalIF":0.0,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148738108","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}