TLDR: While AI tools promise increased efficiency, a new article from The Conversation Canada highlights their potential to erode human trust, creativity, and agency in the workplace. Over-reliance on AI can lead to ‘automation bias,’ diminished self-trust, and even reduced neural connectivity, urging leaders to implement strategies that prioritize critical thinking and ethical human-AI collaboration.
The integration of Artificial Intelligence (AI) into the workplace is rapidly transforming how tasks are performed, with many business leaders prioritizing the efficiency and compliance that generative AI technologies promise. A recent survey indicates that 63 percent of global IT leaders are concerned their companies will fall behind without AI adoption. However, a new analysis from Jordan Loewen-Colón of Queen’s University, Ontario, and Mel Sellick of Arizona State University, published in The Conversation Canada on October 7, 2025, warns that this rush to adopt AI often overlooks its profound impact on workers’ cognition, agency, and cultural norms, potentially eroding trust and creativity.
The authors, researchers in AI, psychology, human-computer interaction, and ethics, express deep concern over the ‘hidden effects and consequences of AI use.’ They argue that AI is not merely a tool for automation and cost-saving; it fundamentally restructures ‘not only what we know, but how we know it.’ As AI becomes more embedded, it will increasingly influence organizational tone, pace, communication style, and decision-making. This necessitates that leaders establish clear boundaries and actively shape the organizational culture around AI integration.
One significant psychological effect identified is ‘automation bias,’ where AI outputs are uncritically accepted as authoritative truths due to their fluent and objective appearance. A study cited revealed that in 40 percent of tasks, knowledge workers accepted AI outputs without any scrutiny, leading to an ‘inflated sense of confidence and a dangerous illusion of competence.’
Another critical concern is the ‘erosion of self-trust.’ Continuous interaction with AI-generated content can cause workers to second-guess their own judgment and become overly reliant on AI guidance. This shift can transform work from generating original ideas to merely approving AI-produced ones, thereby diminishing personal judgment, creativity, and original authorship. Research indicates that users tend to follow AI advice even when it contradicts their own judgment, leading to a decline in confidence and autonomous decision-making. Furthermore, affirming feedback from AI systems, even for incorrect answers, can distort human judgment.
Emerging neurological evidence also supports these concerns. A recent study tracking professionals’ brain activity over four months found that ChatGPT users exhibited 55 percent less neural connectivity compared to those working unassisted. These users also struggled to recall essays they had co-authored moments earlier and showed reduced creative engagement.
To mitigate these risks, Loewen-Colón and Sellick propose several strategies for leaders and managers to foster ‘genuine resilience’ in AI-integrated environments:
1. Systematic Training: Implement training programs that teach interpretive and critical skills, enabling employees to collaborate effectively with AI without over-reliance.
2. Epistemic Awareness: Train individuals to differentiate between fluency and accuracy, encouraging them to question the origin of information rather than passively consuming it. This helps workers become active interpreters of AI outputs.
3. Metacognitive Practices: Encourage self-monitoring, planning, and prompt revision. A study showed that professionals with strong metacognitive practices achieved significantly higher creativity with AI tools.
4. Avoid One-Size-Fits-All: Tailor AI integration by task stages, defining clear roles for AI (drafting, analyzing) and humans (leading, verifying). Incorporate AI-use into responsibility and accountability charts.
5. Culture of Inquiry: Create workplaces that encourage questioning AI outputs, tracking challenges as quality signals, and allocating time for verification. Establish style norms for AI-assisted writing, set confidence thresholds, and specify sign-off procedures for different risk levels.
6. Regular ‘Drift Reviews’: Conduct quarterly reviews to identify shifts in tone, reliance, or bias before they become ingrained in organizational culture.
Also Read:
- Australian Trust in Digital Services Erodes Amidst Rising AI Adoption and Data Security Concerns
- AI’s Impact: White-Collar Job Listings Decline by 20% in South Asia Amidst Rising Demand for AI Skills
The authors conclude that ultimate success in the AI era will not be determined by efficiency alone, but by the ability to critically interpret and assess AI outputs. Companies that balance speed with skepticism and safeguard human judgment as a primary asset are more likely to navigate volatility and thrive.
