Artificial intelligence tools are becoming increasingly common in workplaces, helping employees complete tasks, generate content and improve productivity. However, new research suggests that employees who use large language models (LLMs), such as ChatGPT, may receive lower levels of trust from their supervisors compared with those who complete tasks independently.
The findings come from a peer-reviewed research paper titled Psychological Reactions to Subordinates’ LLM Use at Work, published in The Leadership Quarterly. Researchers Samuel Meyers, Roman Briker and Yochanan E. Bigman conducted six studies, including five incentivised economic experiments and one field study, to examine how employees' use of AI-generated content influences managerial trust.
AI Use Influences Supervisors' Trust in Employees
One experiment examined whether supervisors would accept an employee's report about the time taken to complete a task or independently verify the information.
Employees who completed their work independently received supervisors' trust 68% of the time. However, when employees used AI-generated output, supervisors accepted their reported information without verification just 24% of the time.
The experiment also found that AI-assisted work was viewed more favourably than work copied from another person.
Importantly, supervisors were not evaluating the quality of the completed work in this experiment. Their decisions focused on whether they were willing to trust employees' accounts, suggesting that knowledge of AI involvement itself influenced perceptions of employee trustworthiness.
Lower Trust May Affect Employee Performance Evaluations
The research also identified potential consequences for employee assessments and career development.
Across the studies, lower supervisory trust was associated with poorer perceived job performance, lower assessments of organisational citizenship behaviour and reduced perceptions of employees' promotability.
These findings suggest that managerial attitudes towards AI use may influence how employees are evaluated, even when the research does not establish actual differences in promotions or workplace performance.
Clear Workplace AI Policies Can Reduce the Trust Gap
Organisational policies emerged as an important factor influencing supervisors' reactions to AI-assisted work.
When organisations explicitly encouraged employees to use large language models, the negative effect of AI use on supervisory trust was no longer statistically significant.
However, the negative effect remained when organisations prohibited AI use or had no clearly communicated policy.
The researchers also examined whether task importance and supervisors' previous experience with AI influenced their responses. Neither consistently reduced the negative effect on trust.
Academic Field Study Also Identified Lower Supervisory Trust
A separate field study involving academic supervisors evaluating graduate students' work produced similar findings.
The researchers observed that broader use of large language models in students' work was associated with lower reported trust from their supervisors.
Although this field study provides additional evidence, its academic setting means the findings should not automatically be generalised to every workplace or industry.
What the Findings Mean for HR Teams
The findings highlight a challenge for organisations adopting artificial intelligence. While employers increasingly encourage employees to use AI to improve productivity, unclear expectations may create uncertainty about whether using these tools will affect professional credibility.
This uncertainty could encourage some employees to conceal their AI usage, contributing to what is known as shadow AI the use of AI tools without formal organisational approval or disclosure.
For HR leaders, establishing clear workplace AI policies is therefore important. Organisations should communicate which tasks can involve AI assistance, what level of human review is expected and how employees should disclose AI-generated contributions.
Managers should also receive guidance on evaluating AI-assisted work fairly, focusing on the accuracy, quality and accountability of the final output rather than automatically treating AI usage as a sign of reduced employee capability.
Study Highlights the Importance of Trust in Workplace AI Adoption
The research does not establish that every employee who uses ChatGPT or other AI tools will experience lower managerial trust or reduced promotion prospects.
Instead, it demonstrates how supervisors' perceptions of AI-assisted work can influence workplace evaluations and how explicit organisational encouragement can change those perceptions.
For HR professionals, the findings reinforce the importance of combining AI adoption with clear policies, transparent communication and consistent performance evaluation practices.
