英文_ILO_生成性人工智能与工作_一个精确的全球职业暴露指数_76页_3mb

Generative AI and Jobs: A Refined Global Index of Occupational Exposure Summary

This report presents an updated index of occupational exposure to generative AI (GenAI) developed by the International Labour Organization (ILO). The study refines the 2023 methodology using task-based approaches, incorporating human surveys, expert validation, and AI-assisted scoring to provide more accurate global employment estimates.

Methodology

• Task Assessment: A representative sample of 29,753 tasks from Poland's 6-digit occupational classification was surveyed and validated by experts.
• Automation Scoring: Scores range from 0 (no automation potential) to 1, with adjustments for human input and AI predictions.
• Exposure Framework: Revised into four gradients based on mean scores and standard deviation: Low, Moderate, Significant, and Highest exposure.

Key Findings

• Exposure Levels: Clerical occupations have the highest exposure (Gradient 4). Some professional roles, like financial analysts and web developers, show increased exposure due to AI advancements.
• Global Estimates: ~24% of global employment has GenAI exposure; 3.3% is in the highest exposure category (Gradient 4). Disparities are higher in high-income countries (34%) and among female workers.
• Task Automation: No tasks are fully automatable with current GenAI; impacts focus on job transformation rather than elimination.

Conclusion

GenAI exposure varies by occupation, income, and gender, with potential for both opportunities and challenges. Human-centered policies, social dialogue, and targeted interventions are essential for managing the transition.