This study investigates how algorithms delegate tasks to humans in complex, multi-agent settings using the ride-hailing application Uber as a case study. By analyzing 63 US Uber patents and conducting 22 semi-structured interviews with drivers and passengers, the authors explore human-algorithm interaction beyond simple two-agent configurations. The paper theorizes distributed delegation as a construct encompassing collective hybrid appraisal, distribution, and coordination. Problem Prior research on human-algorithm interaction primarily focuses on one-to-one or many-to-one dyads, assuming delegation is binary or that algorithms act with full autonomy. However, real-world digital platforms operate as complex multi-agent environments where multiple algorithms interact simultaneously with numerous humans. Existing frameworks lack theorization on how collectives of algorithms delegate tasks across groups of humans while remaining dependent on continuous human inputs. Outcome - Conceptualizes distributed delegation as a collective, hybrid, and relational process where multiple algorithms work together while drawing on human data inputs.
- Identifies three key mechanisms of algorithmic delegation in multi-agent settings: collective hybrid appraisal, collective hybrid distribution, and collective hybrid coordination.
- Reveals that delegation functions along a continuum rather than as a complete, binary transfer of rights and responsibilities between algorithms and humans.
- Demonstrates that human involvement—through trace data, choice data, and direct interventions during breakdowns—is necessary for algorithmic agency to operate.

This study employs a Design Science Research (DSR) approach to explore and construct a GDPR-compliant, blockchain-based credential verification system. Focused on the recruitment process, the paper combines stakeholder interviews and system architecture modeling to propose actionable design principles and a prototype platform using self-sovereign identity and zero-knowledge proofs. Problem Traditional credential verification processes in hiring are manual, time-consuming, and susceptible to fraud. Moreover, integrating public blockchain technology with privacy laws like the GDPR creates significant challenges due to the conflict between immutable ledgers and individual privacy rights, such as the right to erasure. Outcome - Identified four major categories of design requirements: trust, automation, usability, and regulation.
- Formulated three core design principles for building GDPR-compliant, self-sovereign identity verification systems.
- Developed and evaluated a prototype system ('SafeRecords') leveraging Polygon ID and zero-knowledge proofs to ensure user data ownership and privacy.

Drawing on Feedback Theory, this study investigates whether, how, and when explanations provided alongside AI feedback enhance human learning and task performance. The research model was evaluated using a randomized between-subjects online experiment with 573 participants completing a image-location matching task, supplemented by focus group discussions with AI experts and users. Problem Although supporting human learning is a primary objective of Explainable Artificial Intelligence (XAI), there is limited understanding of how explanations actually facilitate learning from AI feedback. Specifically, research lacks clarity regarding the underlying theoretical mechanisms and how a user's prior knowledge influences learning and performance outcomes. Outcome - Explanations positively impact learning outcomes for users with less prior knowledge, an effect fully mediated by perceived informativeness.
- Explanations improve task performance directly for users with high prior knowledge.
- For users with low prior knowledge, the positive impact of explanations on task performance is fully mediated by informativeness.

This study investigates how on-screen interruptions affect user performance in mobile settings compared to personal computers (PCs). Drawing on capacity theory of attention, the authors combine a self-report questionnaire with a controlled online experiment to evaluate task-switching and resumption costs across different devices. Problem While mobile devices are ubiquitous, frequent interruptions disrupt cognitive focus during digital tasks. However, little research examines how on-screen interruptions specifically impair tasks performed on mobile devices relative to traditional desktop computers. Outcome - Mobile users suffer higher resumption costs, taking significantly longer to resume a primary task after an interruption compared to PC users.
- Mobile users experience lower task-switching costs because they store less primary-task information in working memory, making them less vulnerable to cognitive conflict during interrupting tasks.
- PC users face greater structural interference and longer completion times on interrupting tasks that share cognitive mechanisms with the primary task.
- Application developers should adapt mobile interfaces by modularizing complex workflows into smaller, independent subtasks to accommodate mobile attention limitations.

This study aims to predict customer complaints and their correlation with customer churn in the telecommunications industry using supervised machine learning models. By examining a dataset of 1,000 mobile service users over a six-month period, the research analyzes call patterns and service error handling to develop proactive retention strategies. Problem Telecommunication companies face significant churn when dissatisfied customers encounter service issues but choose not to lodge direct complaints or experience delayed resolutions. Standard customer service channels often miss silent non-complainers or those voicing grievances through external platforms, placing company reputation and revenue at risk. Outcome - A significant portion of customers encountering service errors (26.7%) choose not to file direct complaints, creating a hidden risk of churn.
- Strong correlation was identified between customer complaints and churn events, proving that customers voicing dissatisfaction are substantially more likely to leave the service.
- Non-churning customers exhibited significantly higher daily call engagement and durations, while churned users showed consistently low call activity across all times.
- Almost 15% of customers filed complaints but did not churn, proving that proactive and timely issue resolution can successfully retain at-risk customers.

This study explores how artificial intelligence is transforming academic knowledge work across teaching, research, and academic service. Utilizing an explorative World Café methodology during the EURAM 2025 Conference, the authors captured lived experiences, reflections, and practices from diverse academic participants. Problem While AI adoption in higher education is rapidly accelerating, existing research remains largely conceptual, technology-centric, and fragmented across isolated academic activities. Consequently, there is a lack of empirical insight into how academics holistically experience AI transformations and adapt their daily work roles. Outcome - Academics actively utilize AI tools like ChatGPT and Copilot primarily for research, followed by teaching and administrative service tasks.
- Key opportunities include automating routine teaching preparation, supporting scholarly writing for non-native English speakers, and streamlining grant application workflows.
- Participants raised major concerns regarding ethical risks, AI hallucinations, the degradation of peer-review quality, and the potential erosion of critical learning and research skills.
- AI impacts are mediated by individual AI literacy, institutional policies, and disciplinary norms, necessitating proactive university governance.
- A research agenda is outlined for Information Systems scholars emphasizing human-AI collaboration, responsible AI design, and the career impacts on early-career academics.

This study examines the prevalence, styles, and sentiment of cyberbullying directed at young women pursuing STEM careers on TikTok in the United States. Combining qualitative content analysis with computational text mining, the author analyzed 33,615 comments across 50 TikTok videos produced by female STEM creators. Problem Women in STEM fields face persistent gender bias and systemic barriers, which are further aggravated by online targeted harassment and cyberbullying. Despite TikTok's growing role in professional networking and youth social interaction, there is limited understanding of how gender-based harassment specifically manifests on the platform. Outcome - Non-cyberbullying comments occurred significantly more frequently than cyberbullying comments on videos created by women in STEM.
- Gendered slurs (34.10%), sexualization (29.64%), and mansplaining (29.04%) were the most predominant cyberbullying tactics, occurring at rates significantly higher than expected.
- Overt harassment tactics like belittlement (2.79%), invalidation (2.06%), discrediting (1.88%), and gatekeeping (0.49%) occurred significantly less frequently than expected.
- Sentiment analysis indicated that neutral responses dominated user interactions, followed by positive expressions, with negative sentiment being the least frequent.

This study examines the extent, style, and tone of cyberbullying directed at young males and females posting about school shootings on TikTok in the United States. Utilizing a mixed-methods approach combining qualitative content analysis, text mining, topic modeling, and sentiment analysis, the author evaluated 57,837 comments across 60 TikTok videos. The research aims to evaluate how national crisis contexts shape peer-to-peer social media interactions and audience engagement. Problem While school shootings generate severe psychological trauma and widespread distress among youth, little is known about how crisis-related stress impacts social media interactions on youth-dominated platforms like TikTok. A key research gap exists regarding gendered differences in online harassment during emergency periods, limiting the effectiveness of digital safety policies and targeted mental health interventions. Outcome - Non-cyberbullying comments occurred significantly more frequently than cyberbullying comments across posts from both male and female youth creators.
- Emotional trauma responses were the predominant cyberbullying behavior, comprising 52.69% of harassment on male posts and 63.97% on female posts, while aggression, sarcasm, and blame occurred less than expected.
- Topic modeling revealed gendered discussion dynamics, where male posts prompted political discourse and threat-focused debates, while female posts elicited supportive, emotionally processing, and healing-oriented conversations.
- Sentiment analysis indicated that posts by young females received a higher proportion of neutral comments and fewer negative reactions compared to posts by young males.

This study examines the prevalence and nature of cyberbullying within gamified TikTok trends targeting U.S. adolescents. Using a mixed-methods design combining qualitative content analysis and computational text mining across 50,755 comments from 60 TikTok videos, the researcher analyzed harassment styles, topic themes, and user sentiment across genders. Problem Gamified social media challenges frame cyberbullying as playful entertainment, making abusive behavior harder for adolescents to identify or report. Research has not adequately explained how platform-specific gamification dynamics on TikTok shape digital peer pressure, cyberbullying styles, and adolescent mental health risks across genders. Outcome - Non-cyberbullying comments comprised over 96% of all interactions, demonstrating the impact of automated platform moderation and bystander intervention.
- Sexualization emerged as the predominant cyberbullying tactic, accounting for 92.46% of male-directed cyberbullying and 89.19% of female-directed cyberbullying.
- Gendered patterns were prominent in topic modeling: male content elicited performance-based roasts, whereas female content attracted appearance-focused rating commentary.
- Sentiment analysis showed that female adolescents received proportionally more neutral audience responses and fewer negative reactions than male peers.

This study utilized a mixed-methods design integrating qualitative content analysis and text mining to explore cyberbullying behaviors on TikTok. The researcher analyzed 83,154 comments across 50 TikTok videos featuring young adults in the United States sharing depressive experiences. The study aimed to evaluate the prevalence, interaction styles, key topics, and overall audience sentiment related to online mental health disclosures. Problem While social media platforms like TikTok have become primary outlets for young adults to disclose mental health struggles, online harassment creates severe barriers to seeking professional help. A critical knowledge gap exists regarding platform-specific cyberbullying dynamics on TikTok and how hostile interactions affect vulnerable youth experiencing depression in the United States. Outcome - Non-cyberbullying comments significantly outnumbered cyberbullying comments, accounting for 92.2% of overall engagement compared to 7.8% cyberbullying interactions.
- Supportive responses were the most common interaction style (53.51%), followed by harmful (17.68%), dismissive (15.91%), sarcastic (8.83%), peer invalidation (3.42%), and blaming (0.65%) comments.
- Sentiment analysis revealed that responses to depression-related content were predominantly neutral, reflecting cautious and non-judgmental user engagement.
- Topic modeling identified five primary thematic areas in user discussions: Emotional Vulnerability, Identity Struggles, Social Connection, Negative Emotions, and Emotional Expression.

This study employs qualitative content analysis and text mining techniques to examine the extent, sentiment, and thematic structure of cyberbullying targeted at adolescent girls' body image on TikTok. Analyzing 80,954 public comments across 50 TikTok videos, the research identifies specific online harassment behaviors and discourse patterns in appearance-centered content. Problem Appearance-related cyberbullying on social media contributes significantly to body dissatisfaction, depression, and disordered eating among adolescent girls in the United States. Federal regulatory agencies and digital safety initiatives require platform-specific evidence to design targeted safety policies and youth mental health interventions. Outcome - Cyberbullying constituted 7.4% of total analyzed interactions, whereas non-cyberbullying comments comprised 92.6%.
- Body shaming was the most prevalent cyberbullying behavior (41.68%), followed by insults (24.78%) and teasing (21.26%).
- Neutral sentiment predominated user responses, followed by positive sentiment, with negative sentiment representing the smallest proportion.
- Topic modeling revealed five primary themes: Appearance Monitoring, Body Commentary, Positive Affirmations, Social Comparison, and Identity Confusion.

This paper proposes a socio-technical research agenda to investigate the underlying drivers and contextual factors of digital gullibility. The authors analyze how user characteristics, information context, and system design interact to influence epistemic vigilance and online decision-making. By establishing ten core propositions across three research gaps, the study provides guidance for creating resilient information systems that mitigate gullible behaviors. Problem Despite the rise of online misinformation, phishing, and digital scams, information systems research traditionally focuses on output symptoms rather than human gullibility as a root cause. Current theoretical models fail to explain why non-malicious users routinely accept untrustworthy online cues and suffer costly consequences. Without understanding these drivers, researchers and designers struggle to build systems that support critical information evaluation. Outcome - Frames digital gullibility as a key antecedent to maladaptive online behaviors rather than focusing solely on misinformation output.
- Integrates socio-technical system principles with concepts like context collapse and network externalities to explain persistent online gullibility.
- Formulates ten research propositions across three gaps: information context, individual seeking behavior, and realized behavior.
- Provides actionable guidance for designing user-centric systems that enhance epistemic vigilance and reduce information verification costs.

This study conducts a scoping review of 33 empirical articles to explore how communities leverage social media platforms to foster digital mental health resilience during natural disasters. Using reflexive thematic analysis, the authors synthesize socio-technical processes across disaster phases to understand how digital environments support psychological adaptation and recovery. Problem While social media is widely recognized for crisis communication and emergency response, prior research has largely overlooked its role in supporting psychological recovery. Furthermore, resilience is often treated as a static endpoint rather than a dynamic process co-constructed through individual, community, and platform interactions. Outcome - Identified seven key socio-technical themes describing how social media fosters mental health resilience across natural disaster phases.
- Revealed that communities actively curate content and self-regulate online spaces to protect users from distress, misinformation, and trauma.
- Demonstrated the importance of vulnerability-sensitive channels and cross-sector coordination among community groups, government agencies, and mental health providers.

This study presents a qualitative assessment of two iterations of an upper-level undergraduate Information Systems course that integrated service learning with Agile software development methodologies. Drawing on qualitative data and stakeholder feedback collected over two years, the paper evaluates how live community partner projects can foster experiential learning and higher-order cognitive skills. Problem While educators and industry demand that Information Systems graduates possess applied, leadership-oriented skillsets, industry partners are frequently hesitant to allow students access to live corporate systems. Integrating service learning with non-profit community clients offers a viable alternative, but it introduces pedagogical tensions regarding instruction time, team dynamics, and project management risks. Outcome - Integrating service learning into Information Systems classrooms increases student motivation and engagement by connecting technical tasks to pro-social community impacts.
- Employing constructivist classroom management alongside Agile methodologies effectively supports higher-level thinking skills, including critical evaluation and self-reflection.
- High team cohesion can be a double-edged sword, occasionally masking individual inefficiencies and inhibiting necessary exchanges with external stakeholders.
- Assessing students on their execution of Agile project management processes rather than purely technical deliverables protects individual grading fairness and encourages risk-taking.

This study presents preliminary insights from an action research case study of a Smart City mobility initiative in the Munich metropolitan region. The researchers developed a 'business ecosystem explorer' software prototype to evaluate various visualization techniques designed to help stakeholders navigate and govern emerging urban mobility platforms. Problem Smart City mobility initiatives involve dynamic, multi-stakeholder ecosystems where firms, public authorities, and infrastructure providers must coordinate across complex digital platforms. Visualizing these nascent, multi-layered platform hubs at a macro level presents significant challenges, making it difficult for stakeholders to identify strategic innovation opportunities and facilitate knowledge sharing. Outcome - Implemented five distinct visualization techniques (Force-directed, Tree Map, Matrix, Radial Network, and Modified Ego-Network layouts) within a software prototype to match diverse stakeholder decision-making needs.
- Demonstrated how macro-level visual analytics can uncover key industry players, platform overlaps, missing service niches, and dominant inter-firm relationships.
- Conceptualized a framework showing how tailored ecosystem visualizations can stimulate vital knowledge flows across policy governance boards, commercial firms, and smart city citizen communities.

This paper investigates how collaborative information systems can be designed to empower domain experts to configure software applications according to their specific needs. Through a qualitative three-year field study of three innovation networks, the authors observe domain experts using a wiki-based collaborative information system and derive design principles for domain expert configuration. Problem Cooperative innovation projects involving multiple partnering firms struggle with managing knowledge, methodical, and relational diversity. Traditional software applications are rigid and cannot easily adapt on-the-fly to the specific, evolving requirements of diverse expert specialists from different organizations. Outcome - Identified key design principles for domain expert configurable software, including iterative modeling, content-context relation, information re-use, function re-use, and context alignment.
- Highlighted the essential role of 'facilitators' as mediating agents who help bridge technical configuration mechanisms with business domain requirements.
- Demonstrated that domain expert configuration provides a flexible, cost-effective method to align IT support with dynamic inter-organizational innovation processes.

This study examines cybersecurity as a complex socially enacted context within organizational boards rather than a simple matter of technical infrastructure. Using the concept of technological frames, it investigates how individual, collective board, and regulatory cognitive models influence board dialogues and shape organizational cyber-resilience. Problem Non-executive board members often lack direct experience with cybersecurity technology, forcing them to rely on indirect proxies and cognitive interpretations. Consequently, cybersecurity dialogues at board levels frequently suffer from incongruent perspectives, resulting in superficial compliance checking, topic avoidance, or political maneuvering. Outcome - Identified three distinct technological frames operating in board discussions: Individual (I-Frame), Board (B-Frame), and Regulatory (R-Frame).
- Revealed dysfunctional dialogue patterns such as topic avoidance and political exploitation that hinder proactive cybersecurity governance.
- Emphasized the necessity of developing conversational guardrails to help boards re-frame discussions toward building true organizational resilience.

This study conducts an exploratory multiple case study based on semi-structured interviews with CEOs of Italian AI firms to investigate how entrepreneurs perceive AI bias within their technological solutions. By examining decision-making processes and organizational contexts, the paper explores how biases are identified, interpreted, and managed in practice. Problem Although conceptual research on AI bias is abundant, empirical studies focusing on how entrepreneurs actually perceive and address bias during AI solution development remain limited. Organizations face challenges in balancing computational accuracy with social and ethical considerations, increasing the risk of operational and strategic failures. Outcome - Revealed two primary interpretations of AI bias held by entrepreneurs: computational (technical) bias and systemic (societal) bias.
- Proposed a 2x2 framework mapping potential negative outcomes and mitigation strategies according to decision levels (operational vs. strategic) and bias types (technical vs. social).
- Showed that computational bias requires iterative algorithmic training and data validation, whereas systemic bias demands targeted stakeholder communication and organizational alignment.
- Emphasized the necessity of adopting a socio-technical approach to foster fair, equitable, and effective AI adoption across different organizational contexts.