Enterprise AI Analysis

Understanding Insider Deviance in the AI Era

Insider deviance refers to the violation of organizational norms by trusted individuals that threaten well-being, involving compromise, manipulation, or unauthorized access to systems. This includes behaviors like fraud, sabotage, and social engineering. Factors such as financial conflicts, job dissatisfaction, social isolation, and resentment can escalate deviant behaviors, particularly in response to AI-driven technologies.

AI-driven surveillance can lead to fears of job replacement and loss of control, fostering anxiety and resentment, which aligns with General Strain Theory (GST) where perceived injustice can lead to deviant behavior as a coping mechanism.

AI Technologies in the Workplace

The integration of AI technologies, such as Performance Monitoring Tools (PMTs) and Automated Decision-Making Systems (ADMSs), is rapidly transforming organizational operations. While these tools enhance efficiency and decision-making, they also introduce new challenges.

AI-driven tools influence hiring, promotions, and disciplinary actions, raising concerns over privacy, technostress, and job insecurity. These systems can lead to increased workplace surveillance, stress, and arbitrary disciplinary actions, potentially causing employees to feel unfairly targeted and increasing psychological distress.

General Strain Theory (GST) in AI Context

General Strain Theory (GST), developed by Agnew (1985, 1992), posits that individuals engage in deviant behavior when exposed to stressors that trigger negative emotions like anger, frustration, or depression. GST identifies three primary forms of strain:

• Failure to achieve positively valued goals.
• Removal of positive stimuli (e.g., job loss, career stagnation).
• Presentation of negative stimuli (e.g., workplace stress, excessive monitoring, unfair treatment).

This study applies GST to explain how AI-driven workplace stressors can increase the likelihood of insider deviance. When AI is poorly managed, strains perceived as unjust or unavoidable can foster negative emotional states, increasing the risk of deviant behavior.

AI-Induced Work Strain

AI-induced work strain (AIWS) is defined as the psychological stress and emotional burden experienced by employees due to AI technology. This includes anxiety over job security, difficulties adapting to AI tools, or cognitive overload from managing complex AI-driven systems. PMTs and ADMSs create psychological and structural pressures when artificial intelligence judgments replace human judgment without consultation, or when employees feel overwhelmed by the rapid pace of skill development requirements.

Hypothesis H1 states: AI-Induced Work Strain Positively Affects Employee Strain. Hypothesis H2 states: AI-Induced Work Strain Positively Affects Insider Deviance.

AI-Induced Workload Changes

AI-induced workload changes (AIWC) refer to the dynamic shifts in job demands, task complexity, and control influenced by AI. While AI technologies may automate routine tasks, they can also increase cognitive demands by requiring employees to interpret and act on AI outputs. If employees feel overloaded or under-challenged, it can lead to psychological strain, emotional distress, and fatigue.

Such changes, especially when perceived as a removal of positive stimuli (like control over tasks or social interactions), can increase the likelihood of deviant behavior as a form of protest or reaction.

Hypothesis H3 states: AI-induced workload Changes Positively Affect Employee Strain. Hypothesis H4 states: AI-induced workload Positively Affect Insider Deviance.

AI-Induced Perceived Inequity

AI-induced perceived inequity (AIPI) refers to employees' perception of unfair treatment concerning AI-driven procedures, outcomes, and interpersonal interactions. This can manifest as unequal access to AI resources, biased performance evaluations, or lack of transparency in decision-making by AI systems. These perceptions can lead to dissatisfaction, mistrust, disengagement, and increased resistance to AI integration.

Perceived inequity represents a failure to achieve positively valued goals, contributing to negative behaviors like cyberloafing, organizational deviance, and retaliation.

Hypothesis H5 states: AI-induced Perceived Inequity Positively Affects Employee Strain.

Research Model & Preliminary Findings

This study employs a quantitative research approach, using a structured survey to gather insights into AI-induced workplace strain, workload changes, perceived inequity, employee strain, and insider deviant behavior. The conceptual framework (Figure 1 in the paper) illustrates how these AI-induced factors contribute to employee strain and ultimately to insider deviance.

A pilot study was conducted to refine the research approach, with 141 responses collected for the full study. Preliminary results suggest a significant correlation between AI-induced workplace strains and an increase in insider deviant behaviors, with negative emotional responses like frustration, anxiety, resentment, and distrust playing a key role.