The Story

Anthropic has published new research examining how artificial intelligence may reshape labor markets. The study analyzes real-world usage patterns of the Claude AI system and compares them with occupational tasks to estimate how exposed different jobs may be to AI automation.

How the study measures AI exposure

• Observed exposure metric: The research uses a metric called “observed exposure,” which compares what AI systems are capable of doing with how people are currently using them in practice.
• Task-level analysis: Instead of measuring entire occupations, the study focuses on the individual tasks within jobs that AI tools can already assist with or automate.
• Real usage data: The analysis is based on millions of Claude interactions, providing insight into how AI is actually being used in professional workflows.

Jobs most exposed to AI tools

• Software and programming: Computer programmers rank highest in exposure, with roughly 75% of tasks potentially assisted or automated by AI systems.
• Customer service and data entry: These roles follow closely, with around 67% task exposure due to the structured nature of their workflows.
• Hands-on professions: Roughly one-third of the workforce currently has minimal AI exposure, particularly in physical or service roles such as cooks, bartenders, and lifeguards.

Early signals in the labor market

The study notes that large-scale unemployment directly tied to AI has not yet appeared since ChatGPT’s launch in 2022. However, the data suggests early shifts in hiring patterns. Recruitment into AI-exposed occupations among workers aged 22 to 25 has reportedly declined by around 14% during the same period.

Why it matters

The findings add to a growing debate about AI’s impact on the global workforce. While automation has not yet caused widespread job losses, the study suggests that younger workers entering knowledge-work fields may already be encountering tighter hiring conditions as AI tools become integrated into everyday workflows.

Anthropic CEO Dario Amodei has repeatedly warned that AI could significantly reshape knowledge work over the coming decade. The report highlights how the shift may begin gradually through hiring changes and task automation rather than immediate job displacement.