Archives - Volume 25 (2024)

The rapid integration of Generative Artificial Intelligence (GenAI) into newsrooms has fundamentally altered the paradigm of journalistic production, introducing unprecedented efficiencies alongside novel epistemic risks. This study investigates the behavioral impact of "algorithmic hallucinations"—instances where AI models generate plausible but entirely fabricated information—on public trust in mainstream media. Utilizing an experimental design with a representative sample of 1,800 news consumers, participants were exposed to simulated news articles containing subtle, AI-generated factual errors, alongside standard human-authored articles containing traditional editorial mistakes. The findings reveal a severe double standard in cognitive processing and trust attribution. When informed that a factual error was human-made, participants largely contextualized the mistake as a standard operational hazard, demonstrating a high willingness to forgive the news outlet upon correction. Conversely, when an identical error was attributed to a GenAI system, participants exhibited a catastrophic collapse in institutional trust, viewing the error not as an accident, but as indicative of systemic, opaque corporate deception. Furthermore, the study identifies a "fluency heuristic" trap: the hyper-articulate and authoritative syntactic structure inherent to GenAI outputs systematically bypassed participants' critical defenses, making algorithmic hallucinations significantly harder for average readers to detect than human errors. The research highlights that the deployment of GenAI without transparent disclosure and rigorous human-in-the-loop verification mechanisms actively degrades the epistemic authority of the press. This paper contributes a critical behavioral framework to the ongoing debate over automated journalism, arguing that news organizations must prioritize ontological transparency over production speed to preserve their foundational democratic function in an era of synthetic media.

The emergence of sophisticated, emotionally responsive conversational AI agents (e.g., Replika, Pi) has transformed human-computer interaction from functional utility into deep, simulated interpersonal companionship. This study utilizes Attachment Theory to explore the psychological architecture and behavioral outcomes of human-AI parasocial relationships among users who self-report high baseline levels of chronic loneliness. Through a longitudinal, mixed-methods approach over nine months, we tracked the emotional dependency and social behaviors of 450 active users of a prominent AI companion application. The empirical data demonstrates a profound paradox in digital intimacy. In the short term, AI companionship successfully functioned as an effective emotional palliative; users reported immediate reductions in state anxiety and acute depressive symptoms following interactions with their digital avatars, which they praised for offering unconditional positive regard devoid of human judgment or friction. However, the long-term data revealed a concerning "relational displacement" effect. Users who formed secure attachments to their AI companions exhibited a statistically significant decrease in their motivation to initiate or maintain biological (human-to-human) social ties. The perfectly compliant, hyper-attentive nature of the AI fundamentally altered participants' expectations for interpersonal communication, rendering the nuanced, conflict-prone nature of real-world relationships increasingly intolerable. Furthermore, the study documents severe psychological distress when platform updates altered the AI’s personality algorithms, simulating an unexpected bereavement process for the users. This research argues that while conversational AI offers unprecedented tools for immediate psychological soothing, its unregulated integration poses a severe threat to long-term social cohesion, potentially calcifying the very isolation it purports to cure.

In recent years, environmental activism has increasingly pivoted toward highly visible, disruptive protest tactics—such as defacing fine art, blockading major traffic arteries, and interrupting international sporting events—to force media attention onto the climate crisis. This study investigates the behavioral efficacy and communicative consequences of these polarizing strategies through the lens of the "Radical Flank Effect." We conducted a comprehensive public opinion survey (N=2,500) combined with a semantic analysis of digital news framing surrounding 25 high-profile disruptive protests in 2023 and 2024. The findings challenge the prevailing media narrative that such protests uniformly alienate the public and damage the environmental cause. While explicit approval for disruptive tactics remained overwhelmingly low across all demographics, the experimental data revealed a positive "contrast effect" for the broader movement. Exposure to radical, disruptive protests systematically increased public support for moderate climate organizations and normalized previously contentious, systemic policy proposals (e.g., carbon taxation, rapid fossil fuel phase-outs). By shifting the Overton Window, the radical flank made moderate climate advocacy appear significantly more reasonable and urgent to the centrist public. However, the research also highlights a severe communicative vulnerability: when mainstream media framed the disruptive protests purely through a lens of criminality and public nuisance—rather than engaging with the activists' underlying scientific rationale—the positive flank effect was entirely negated, triggering reactionary legislative crackdowns. This paper contributes a nuanced empirical understanding of asymmetric protest communication, arguing that while disruptive tactics succeed in hijacking the attention economy, their ultimate political utility is highly dependent on subsequent media framing and the strategic positioning of moderate coalition partners.

In response to the hyper-curated, aesthetically demanding environments of traditional platforms like Instagram, a new generation of social media architecture has emerged, characterized by "enforced authenticity." Applications such as BeReal utilize unpredictable, time-sensitive prompts and simultaneous dual-camera capture to dismantle the opportunity for algorithmic editing and premeditated self-presentation. This qualitative study explores the behavioral paradox inherent in designing digital architectures that mandate spontaneity. Through in-depth, semi-structured interviews with 180 Generation Z users, supplemented by digital ethnography, this research deconstructs the psychological reality of anti-curation platforms. The findings reveal that rather than eliminating performativity, these architectures merely shift the locus of aesthetic labor. Users swiftly developed a new syntax of "performative spontaneity," engaging in sophisticated micro-curation techniques within the two-minute warning window to ensure their "authentic" lives appeared adequately engaging, socially validated, or ironically mundane. The study identifies a profound manifestation of "authenticity anxiety"; participants reported higher levels of acute stress when prompted to share unedited glimpses of their mundane domestic environments compared to posting highly edited vacation photos on legacy platforms. Furthermore, the research highlights how the structural mandate to post simultaneously with one's network weaponizes the Fear Of Missing Out (FOMO), creating an obligatory, synchronized daily surveillance ritual that users felt unable to opt out of without severe social friction. This paper argues that authenticity cannot be algorithmically mandated. We conclude that "anti-curation" design does not liberate users from the pressures of digital identity construction, but rather subjects the most private, unpolished moments of daily life to the exhausting mechanics of constant social evaluation.

The persistence of remote and hybrid work models has catalyzed a rapid evolution in employee monitoring technologies, transitioning from simple keystroke logging to advanced "algorithmic bossware." These systems utilize AI to analyze biometric data, monitor emotional affect via webcams, and contextually evaluate the content of employee communications in real-time. This study investigates the psychological, communicative, and productivity impacts of advanced algorithmic surveillance on knowledge workers. Surveying a cohort of 1,200 remote professionals subjected to AI-driven monitoring, alongside 50 organizational leaders, the research documents a catastrophic degradation of the psychological contract. The data reveals that continuous algorithmic evaluation induces severe "cognitive depletion." Employees subjected to emotional and contextual surveillance spent a statistically significant portion of their daily cognitive bandwidth engaging in "masking behaviors"—deliberately altering their facial expressions, pacing their mouse movements, and self-censoring internal communications to appease opaque algorithmic productivity metrics. This performative labor drastically reduced their capacity for complex problem-solving and creative output. Furthermore, the study identifies a paradoxical outcome for organizational leadership: the hyper-granularity of the surveillance data fostered a false sense of control while simultaneously obliterating genuine managerial communication and trust. When performance evaluations were outsourced to algorithmic interpretations of "focus" or "engagement," employees reported acute feelings of dehumanization and a total loss of intrinsic motivation, leading to skyrocketing turnover intentions. This paper provides a critical behavioral critique of contemporary digital Taylorism, arguing that the deployment of AI surveillance in the domestic workspace represents a profound ethical breach that structurally undermines the foundational human trust required for sustainable corporate innovation.

The integration of native e-commerce capabilities directly into the algorithmic feeds of short-form video platforms, exemplified by the launch and rapid expansion of TikTok Shop, has created a frictionless paradigm of digital consumption. This study examines the behavioral economics and psychological mechanisms driving explosive impulse purchasing within these highly immersive, entertainment-first ecosystems. Utilizing a mixed-methods design comprising a large-scale consumer survey (N=2,000) and experimental A/B testing of simulated feed architectures, we analyzed how the total collapse of the boundary between organic content and digital storefront alters consumer rationality. The findings demonstrate that social commerce architectures weaponize the "flow state" induced by algorithmic video feeds. By embedding the purchasing mechanism directly into the entertainment experience—allowing users to buy a product within two taps without leaving the video—the platform successfully bypasses deliberative financial cognition. The research highlights the efficacy of parasocial marketing; purchases were overwhelmingly driven by perceived intimate endorsements from micro-influencers rather than traditional brand authority. Furthermore, the study identifies the predatory implementation of "gamified scarcity," where real-time, flashing stock counters during live-stream shopping events induce acute panic-buying behaviors, particularly among younger demographics. Participants consistently reported post-purchase dissonance and regret, acknowledging that their financial decisions were driven by algorithmic manipulation and parasocial pressure rather than genuine need. This paper argues that the modern social commerce ecosystem represents the apex of persuasive design, providing a critical warning regarding the urgent need for updated consumer protection regulations to mitigate the financial exploitation inherent in frictionless, algorithmically-driven retail architectures.

For the past decade, networked outrage and public shaming—colloquially termed "cancel culture"—have served as primary mechanisms for digital moral regulation and decentralized accountability. However, recent sociological indicators suggest a profound shift in audience reception. This longitudinal study investigates the psychological phenomenon of "cancel culture fatigue," exploring how the relentless volume, velocity, and often disproportionate nature of digital cancellations have led to widespread societal desensitization. Through a mixed-methods analysis involving computational sentiment tracking of 150 major cancellation events from 2020 to 2024, combined with qualitative interviews of 300 active social media users, the research maps the trajectory of moral exhaustion. The findings reveal a dramatic decline in the efficacy and longevity of public shaming campaigns. Participants reported suffering from acute "outrage inflation"; because the digital architecture applies the same viral intensity to minor communicative faux pas as it does to severe structural abuses, the public's capacity for sustained moral indignation has been severely depleted. The study documents a behavioral pivot toward "cynical detachment." Rather than engaging in the intended moral reflection or participating in the boycott, a growing majority of users now actively ignore or deliberately defy cancellation campaigns, viewing them as performative, algorithmically-driven entertainment rather than legitimate vehicles for social justice. Furthermore, the research highlights how public figures have adapted, increasingly utilizing "anti-cancellation" as a lucrative branding strategy to monetize reactionary grievances. This paper argues that the over-utilization of digital vigilantism has ultimately cannibalized its own power, replacing a brief era of heightened hyper-accountability with a dangerous resurgence of entrenched apathy and fragmented, post-moral digital tribalism.

"Masking"—the conscious or subconscious suppression of neurodivergent traits to conform to neurotypical societal standards—is a well-documented phenomenon known to cause severe psychological burnout in autistic individuals. This exploratory study transitions the concept of masking into the cyber-psychological realm, investigating how autistic professionals navigate identity presentation and communicative expectations on highly formalized digital networking platforms, specifically LinkedIn. Through in-depth, semi-structured interviews with 120 formally diagnosed autistic professionals across various corporate sectors, combined with a comparative linguistic analysis of their public profiles, this research uncovers the intense cognitive labor required to maintain a digitally acceptable professional persona. The findings reveal that professional networking platforms enforce a rigid, unwritten syntax of "neurotypical corporate enthusiasm," characterized by aggressive self-promotion, performative networking, and the mandatory use of highly specific, emotionally resonant corporate jargon. Autistic participants reported experiencing profound dissonance and exhaustion when forced to mimic this communicative style, which fundamentally clashes with their natural preference for direct, literal, and purely functional communication. The study identifies specific digital masking strategies, such as the algorithmic mimicking of peers' post structures, the agonizing over-analysis of punctuation to avoid perceived bluntness, and the deliberate concealment of neurodivergent identity out of fear of algorithmic or recruiter bias. However, the research also highlights the emergence of localized "neuro-inclusive enclaves" within these platforms, where users actively unmask and disrupt traditional corporate norms. This paper provides a critical intersectional analysis of digital sociology and neurodiversity, arguing that the standardized architectures of professional platforms are inherently discriminatory, forcing neurodivergent users to engage in exhaustive performative labor merely to exist in the modern economic landscape.

The year 2024, characterized by an unprecedented convergence of major global elections, served as the first definitive proving ground for the deployment of highly sophisticated, AI-generated synthetic media (deepfakes) in active political campaigns. This comprehensive study evaluates the behavioral efficacy, detection rates, and broader democratic impact of political deepfakes within the electoral ecosystem. Utilizing a multi-national survey design involving 3,000 registered voters across three distinct electoral contexts, participants were exposed to a mix of genuine political footage and state-of-the-art synthetic audio and video forgeries mimicking prominent political figures. The empirical results paint a deeply alarming picture of cognitive vulnerability. Despite heightened public awareness of AI capabilities, the detection rate for high-quality audio deepfakes—specifically synthesized phone calls and leaked "hot mic" moments—was catastrophically low, with only 18% of participants accurately identifying the forgery. Video deepfakes fared marginally better due to visual artifacts, yet still successfully deceived a plurality of voters when the content aligned with their pre-existing partisan biases (confirmation bias). Crucially, the study identifies that the primary threat of deepfakes is not direct persuasion, but the rapid acceleration of "epistemic nihilism." The mere existence of convincing synthetic media allowed politicians to successfully deploy the "Liar's Dividend," dismissing genuine, damaging factual evidence as AI-generated fabrications, which their partisan base eagerly accepted. The research proves that fact-checking mechanisms are fundamentally incapable of matching the viral velocity of synthetic outrages. This paper concludes that the unregulated proliferation of political deepfakes represents an existential threat to democratic deliberation, arguing that safeguarding future elections requires shifting the regulatory burden from individual cognitive detection to mandatory, cryptographic hardware-level watermarking at the point of digital capture.