Unaccountable Delegation, Fading Skills: Mapping the Risks of Workplace AI Agents

2026-08-09

AI Summary

This study addresses the limitations of existing AI risk taxonomies, which overlook job-specific contexts and thus fail to support organizations in effectively managing human–AI collaboration risks. To bridge this gap, the authors propose an analytical framework integrating agent, objective, and environment dimensions. Leveraging 2,078 tasks from the O*NET database, they generate 8,356 risk scenarios and combine structured prompt engineering, large language model evaluations, and empirical worker surveys to develop the first occupation-specific taxonomy of workplace AI agent risks. The resulting taxonomy identifies 15 distinct risk categories, revealing pervasive yet often hidden risks—such as skill atrophy—in augmentation-oriented deployments. Crucially, the findings indicate that these risks stem primarily from the design of human–AI collaboration rather than from AI capabilities per se. Empirical validation demonstrates that the proposed taxonomy significantly outperforms existing approaches in terms of usability and user preference.

Abstract

To anticipate socio-technical risks from AI agents, organizations need taxonomies to classify them. However, existing AI risk taxonomies focus on broad risks and do not capture job-specific risks introduced by agents. To address this gap, we make three main contributions. First, we developed a multi-layer framework from a literature review of AI agents. The framework models three core components and their interactions: agents, goals, and environment. Second, we embedded this framework in a structured prompt and applied it to descriptions of 2,078 job tasks from the O*NET database, producing 8,356 risk scenarios labeled by severity and deployment mode (automation or augmentation). We validated these scenarios with 45 workers across 10 job roles and an independent LLM judge, confirming their plausibility and alignment with job tasks. Finally, we extended an existing taxonomy to create a 15-category taxonomy of workplace AI agent risks that covers all our risk scenarios. Our analysis highlights four findings. First, augmentation is not inherently safe because overreliance on agents can gradually erode workers' skills and oversight. Second, Erroneous Agent Actions accounts for the largest share of risk scenarios and has the highest concentration of severe risks. Many arise at the human-agent boundary. Third, automation is associated mainly with organizational risks, while augmentation is associated mainly with risks to workers. Fourth, workers found our taxonomy easier to use for a risk classification task than two other taxonomies and preferred it in 64% of non-tied comparisons with a recent generative AI risk taxonomy. These findings show that workplace AI agent risks do not arise from agents alone; they also depend on how people work with agents and how agents are deployed. Safer workplaces require not only safer agents but also carefully designed human-AI agent collaboration.

Problem

• AI agents
• workplace risks
• risk taxonomy
• job-specific risks
• human-AI collaboration

Innovation

• AI agent risk taxonomy
• human-AI collaboration
• structured prompting
• job-specific risk assessment
• skill erosion