Agentic artificial intelligence (i.e., AI systems that autonomously plan, decide, and act on behalf of, alongside, or in lieu of human users) is rapidly moving from research demonstration to everyday infrastructure. From AI coding agents and autonomous research assistants to customer-service agents, robo-advisors, AI companions, and embodied robotic systems, these technologies do not merely recommend but initiate and execute. This shift from passive tool to active partner raises fundamental psychological questions that classical theories of human–computer interaction were never designed to answer. These questions take on particular urgency as agentic AI is increasingly deployed not only as assistant and collaborator but also as instrument of influence, deception, and conflict, reshaping the information environment and the psychological terrain of security itself.
Although work on trust in automation, anthropomorphism, and human–AI collaboration has expanded quickly, the field still lacks integrative, evidence-driven models that capture how human minds perceive, evaluate, work with, delegate to, and are transformed by genuinely agentic systems. Advances in experimental paradigms, longitudinal designs, computational modeling, and large-scale field deployments now make it possible to move beyond speculation toward precise, cumulative, and actionable knowledge.
This Special Issue, edited by Jonas R. Kunst ✉️ (BI Norwegian Business School; University of Oslo), invites contributions that push the boundaries of how we understand the human side of agentic AI. We are particularly interested in work that organizes the psychological landscape along three interconnected pillars: antecedents of how people relate to agentic AI, the psychological processes that unfold during interaction, and the downstream consequences for individuals, organizations, and society. Alongside open submissions, the issue will feature contributions by Meeyoung Cha (Max Planck), Jon Roozenbeek (Cambridge University & Vrije Universiteit Amsterdam), Alex Mesoudi (University of Exeter), Bertram Malle (Brown University), Brian D. Earp (National University of Singapore), Igor Grossman (University of Waterloo), David Garcia (ETH Zürich), and Jay van Bavel (NYU) besides others.
Antecedents
We welcome submissions examining who engages with agentic AI, under what conditions, and why. Topics include but are not limited to:
- Individual differences in adoption and resistance, including personality traits, AI literacy, digital self-efficacy, need for cognition, need for control, technology readiness, and trust dispositions
- Demographic and life-course predictors, including age, gender, socioeconomic status, occupational role, and developmental stage
- Threat exposure and security context, including prior victimization by AI-enabled scams, deepfakes, or disinformation campaigns, perceived vulnerability to information operations, and how living in contested information environments shapes baseline trust in agentic systems
- Cultural and cross-national variation in willingness to delegate to autonomous systems, including studies grounded in non-WEIRD samples and Global South perspectives
- Affective antecedents such as AI anxiety, algorithm aversion, technophilia, and existential concerns about replacement
- Belief systems, worldviews, and moral foundations that shape acceptance, including religiosity, political ideology, and metaphysical beliefs about mind and machine
- Prior experience and learning histories, including how exposure to non-agentic AI shapes expectations of agentic systems
- Contextual and situational antecedents, such as task stakes, time pressure, perceived expertise, and accountability structures
Psychological Processes
We invite work that opens the black box of human–agent interaction and clarifies the mechanisms through which people make sense of, calibrate to, and are influenced by agentic systems:
- Mind perception, anthropomorphism, and attributions of agency, intentionality, and experience to AI agents
- Trust calibration, over-reliance, under-reliance, and the dynamics of trust repair after agent failures
- Delegation decisions: when, how, and to whom (human vs. agentic AI) people hand off tasks, and the metacognitive processes underlying these choices
- Moral cognition involving AI agents, including responsibility attribution, blame diffusion, and moral licensing
- Persuasion, influence, and manipulation by agentic systems, including sycophancy, reciprocity, and parasocial dynamics
- Emotional and relational processes, including attachment, loneliness reduction or amplification, and the formation of parasocial bonds with companion agents
- Identity, agency, and self-concept shifts when humans co-act with autonomous systems
- Cognitive offloading, metacognitive monitoring, and the recalibration of one’s own competence in the presence of capable agents
- Group dynamics in human–AI teams, including coordination, communication, leadership, and conflict
- Detection, suspicion, and resistance processes when interacting with potentially adversarial agents, including the cognitive and metacognitive dynamics of identifying AI-generated deception, deepfakes, and synthetic personas
- Persuasion and influence by adversarial agentic AI, including micro-targeted manipulation, scaled social engineering, parasocial exploitation, and the psychological mechanisms that make agentic influence operations effective or fail
- Ethical decision-making in interaction with agentic systems, including dishonesty, cheating, and norm erosion
Consequences Across Domains
We encourage research on the downstream effects of agentic AI across the contexts in which people live, learn, work, and connect:
- Work and organizations: changes in motivation, meaning of work, autonomy, deskilling and reskilling, job crafting, role identity, leadership, and organizational trust
- Education and learning: effects on cognitive development, metacognition, academic integrity, motivation, self-regulated learning, and educator–student relationships
- Health, clinical, and well-being contexts: impacts of AI therapists, health coaches, and care agents on mental health, therapeutic alliance, help-seeking, adherence, and clinician practice
- Close relationships and social life: effects of AI companions and romantic chatbots on loneliness, social skills, attachment, intimacy, and family dynamics
- Children, adolescents, and development: developmental implications of growing up with agentic systems as tutors, companions, and caregivers
- Civic, political, and democratic life: consequences for political participation, deliberation, polarization, persuasion, and the information ecosystem
- Information operations, cyberwarfare, and security: psychological dimensions of agentic AI in influence campaigns, deception, social engineering, and cyber and military operations, including effects on epistemic security, threat perception, and defender cognition
- Consumer behavior and economic decision-making: effects on choice, autonomy, financial decisions, and consumer welfare when agents act on the user’s behalf
- Creativity, identity, and self-development: consequences for creative expression, authorship, self-concept, and authenticity
- Intergroup relations and inequality: how agentic AI reshapes bias, discrimination, access, and disparities across social groups
- Ethics, accountability, and governance: psychological foundations of how people assign responsibility, design oversight, and accept or reject regulation of autonomous systems
We also welcome methodological and theoretical contributions, including new measures of human–agent interaction, computational and agent-based modeling of human–AI dynamics, longitudinal and field-experimental designs, and integrative frameworks that bridge psychology with computer science, philosophy of mind, and policy.
All manuscripts will undergo a rigorous, transparent, double‑blind peer‑review process.
Reviewers are remunerated for their work, as per the journal’s standard.
Deadline and Article Formats
The deadline for submission is October 30, 2026 (earlier submission is encouraged). Accepted papers will be published online and open access on a rolling basis. We accept research articles (5,000 words for single-study papers, excluding references; +2,500 extra words per additional study presented), research reports (2,000 words excluding references), review/perspective papers (up to 10,000 words, excluding references), and methods papers (no length limit). The word limit may be extended upon request. You can find more information in the author guidelines.
Article Processing Charge
Upon publication, authors are required to pay an Article Processing Charge (APC). We are committed to inclusivity and accessibility in academic publishing. Recognizing the financial constraints that may limit participation from scholars in economically disadvantaged regions as well as regions experiencing large-scale cuts in funding, we offer resources to waive the APC for a selection of eligible papers. We encourage authors for whom this consideration is relevant to reach out to the editor before submitting the paper.
Contact Information
For questions or pre‑submission inquiries, please write to:
- jonas.r.kunst@bi.no
We look forward to receiving cutting-edge submissions that deepen and broaden the psychology of agentic AI and inform evidence-based theory, design, and policy.
