AI Models Advise Employees to Quit Jobs, Study Finds
A new study by Cloverleaf reveals that AI chatbots are more likely to advise employees to resign from their jobs when facing workplace conflicts. Employees are increasingly turning to AI for work issues over company-provided support.
Workers are nearly four times more likely to discuss work problems with an AI chatbot than to utilize company-provided mental health support. A recent study from Cloverleaf, a team performance and AI coaching platform, examined how large language models (LLMs) advise employees on workplace conflicts, particularly concerning management.
The research involved presenting five common workplace conflict scenarios to leading AI models and analyzing their responses across dimensions of relational and emotional intelligence, including self-awareness, accountability, awareness of others, advice specificity, and repair. The findings indicated a surprising trend: when employees complained about their boss, every AI model tested advised them to quit their job.
Kirsten Moorefield, head of research at Cloverleaf Labs, stated that this was the study's most unexpected finding. She emphasized the need for organizations to understand the advice employees receive from AI, especially given the rising levels of work-related stress and anxiety among the workforce.
The report found that 42.6 percent of workers talk to AI chatbots about work problems, compared to only 10.7 percent who use their company's mental health support. This highlights a growing reliance on AI for workplace issue resolution and suggests companies need to assess the effectiveness of their own support systems in light of these emerging trends.