人工智慧系統導入職場之工作健康安全議題

Occupational Health and Safety Issues Derived from the Introduction of Artificial Intelligence Systems to Workplace

人工智慧（Artificial Intelligence, AI）作為重要驅動力，雖引領產業跨越智慧時代的門檻，然而AI科技應用於職場生產和勞工管理所衍生的職業健康安全（Occupational Safety and Health, OSH）議題既新穎又值得關注。因此，本文透過美國、歐盟及澳洲相關單位及國際組織之官方網站及文獻搜尋，探討國際上新近發展的AI職場應用及相關因應政策方向，希望透過統整提供我國參考。本文發現，AI系統導入職場雖提高安全衛生管理的效率與完整性，但也模糊了個人隱私和勞工自主權的界線，同時增加勞工在面對組織或作業型態及工作量改變的社會心理壓力，又或者反而增加重複性動作而導致人因性危害等問題。針對AI導入的風險，美國、歐盟及澳洲紛紛建立AI使用的基本規範，並提出預防性管理的指引或建議。澳洲新南威爾斯州政府（NSW Government）甚至發布AI工作健康與安全計分卡（Artificial Intelligence WHS Scorecard，簡稱AI WHS計分卡）作為事業單位風險評估的工具，指導雇主站在勞工的角度去評估工作場所使用AI系統相關的OSH風險。建議我國可參考歐盟以風險為基礎的分級管理原則，將涉及勞工健康安全危害之AI系統列為高風險管理等級，並遵循以人為本的原則，建立AI應用的勞工健康安全規範。也可以參考澳洲AI WHS計分卡的風險評估做法，針對AI應用於勞工管理或智慧機器人與勞工協作的職場，發展事業單位AI系統風險評估的工具，並透過發展產業指引及教育訓練，引導事業單位更安全且負責任的導入及應用AI系統，以兼顧智慧生產與管理及勞工工作健康安全。

Artificial Intelligence (AI) has emerged as a significant driving force, leading industries into the smart era. However, the application of AI technology in workplace production and labor management has introduced new and noteworthy occupational safety and health (OSH) issues. This article explores the latest developments in workplace AI applications and relevant policies internationally by reviewing the official websites and literature from OSH-related units in the United States, the European Union (EU), Australia, and international organizations, aiming to provide a reference for Taiwan. Thefindings reveal that while the introduction of AI in the workplace improves the efficiency and integrity of safety and health management, it also blurs the lines between personal privacy and worker autonomy. Additionally, AI introduces new psychosocial stressors for workers due to changes in organizational structures, operational patterns, and workloads. It may also increase ergonomic hazards through repetitive motions. To address the risks associated with AI, the United States, the EU, and Australia have established basic regulations for AI use and issued preventive management guidelines/suggestions. The New South Wales Government in Australia has even released the AI Work Health and Safety (AI WHS) scorecard as a risk assessment tool for enterprises, guiding employers to evaluate OSH risks associated with AI from the workers' perspective. It is recommended that Taiwan consider adopting the EU's risk-based tiered management principles, classifying AI systems that pose safety and health hazards to workers as high-risk, and establishing AI-related OSH regulations grounded in human-centered principles. Furthermore, Taiwan could refer to risk assessment practices of Australia’s AI WHS scorecard, developing tools for risk assessment of AI applications in labor management or collaborative workplaces with smart robots. This approach should be supported by industry guidelines and educational training to ensure safe and responsible integration of AI systems, balancing smart production/management and worker health and safety.