Anthropic unveils AI's unreplaceable jobs

Understanding AI's Impact on the Workforce

Artificial intelligence (AI) is reshaping the job market in ways that are both exciting and concerning. While many people have their own opinions on which jobs will be most affected by AI, companies like Anthropic are now providing concrete data to better understand this shift. Their research introduces a new metric called "observed exposure," which tracks how AI is being used in real-world workplaces rather than just what it could theoretically do. This distinction is crucial, as it reveals a significant gap between AI's potential and its actual impact.

The Observed Exposure Metric

Anthropic’s team developed the "observed exposure" metric by combining three key sources of information: the O*NET database, which covers roughly 800 U.S. occupations; Claude’s own usage logs; and a 2023 academic framework that evaluates whether AI can cut a task's completion time in half. Each job receives a coverage score, with higher scores indicating that AI is already handling a substantial portion of the job's tasks. A score of zero means that AI has not shown up in the data at all.

This approach provides a more accurate picture of AI's role in the workforce. For instance, while AI could theoretically handle 90% of tasks in office and administrative roles, the observed usage only covers about a third of computer and math jobs, which are already the most penetrated category. This gap highlights the complexity of AI's impact and suggests that many jobs remain relatively safe for now.

Jobs Feeling the Pressure from AI

Some professions are experiencing more direct pressure from AI than others. Computer programmers lead the list with 75% task coverage. Claude is being used heavily for coding, and this usage leans toward full automation rather than simply helping programmers work faster. Customer service representatives come in second, as their core tasks increasingly appear in first-party API traffic, indicating that companies are quietly replacing human agents with AI pipelines. Data-entry workers are also feeling the impact, with 67% coverage due to the efficiency and cost-effectiveness of AI in reading documents and entering data.

Other high-exposure occupations include: - Financial analysts, whose modeling and number-crunching work is heavily covered. - Office administrators, facing 90% theoretical exposure, even if real adoption still lags. - Computer and math workers broadly, where observed exposure sits at 33% and climbing.

The Bureau of Labor Statistics supports these findings, showing that for every 10 percentage point increase in a job's AI coverage score, projected growth for that role drops by 0.6 percentage points. While this may not be catastrophic, the direction is clear.

Jobs AI Cannot Touch

Approximately 30% of U.S. workers score a flat zero on the AI coverage scale. These roles involve tasks that require physical presence, sensory judgment, and the ability to read the room in real time—areas where AI currently falls short. Examples include:

Cooks, who rely on knife skills, tasting, and plating judgment.
Motorcycle mechanics, who diagnose engines through hands-on inspection.
Lifeguards, whose job involves scanning water and executing physical rescues.
Bartenders, who read crowds and social dynamics in real time.
Dishwashers and dressing-room attendants, handling wet, physical, and unpredictable tasks.
Agricultural workers pruning trees and operating farm machinery outdoors.
Courtroom lawyers, whose work demands physical presence and live advocacy.

These jobs are expected to see steady growth, particularly in healthcare, where demand for nurses, therapists, and care workers is outpacing anything AI can displace.

Who Faces the Greatest AI Threat?

The research also highlights an unexpected trend: the workers most at risk from AI are not the ones typically associated with low-wage jobs. Instead, they tend to be older, more educated, female, and significantly better paid, earning about 47% more than their zero-exposure counterparts. This shift marks a departure from previous automation waves, which primarily affected lower-wage workers.

Despite this, there is no unemployment crisis to report yet. The study found no measurable rise in joblessness among high-exposure workers since ChatGPT launched. However, the impact is beginning to show in hiring practices. Among workers aged 22 to 25, the monthly job-finding rate in high-exposure occupations has fallen roughly 14% since ChatGPT's arrival. While this drop is barely statistically significant, it aligns with findings from other researchers tracking ADP payroll data, which suggest that young people trying to enter exposed fields are finding fewer opportunities.

Implications for Markets and Policy

Investors are already adjusting their strategies in response to growing concerns about AI's impact on white-collar jobs. Health care and utilities are seeing increased allocations from institutional money, while software-heavy tech sectors face challenges. The Nasdaq slipped about 1.2% after the report circulated, reflecting these shifts.

Policymakers are also taking action, albeit slowly. Retraining programs targeting trades and care fields are gaining traction, with proposals including tax credits for physical-trade apprenticeships, immigration pathways to fill care-sector gaps, and wage subsidies for frontline jobs where AI pressure is low.

Anthropic's researchers emphasize that this is just the beginning of their work. They plan to continue updating the coverage measures as usage data evolves and closely monitor whether the youth hiring slowdown deepens.

Conclusion

While AI-fueled mass displacement has not yet arrived, the early signals are pointing in one direction. Anyone paying attention to where younger workers aren't getting hired should take note. The future of work remains uncertain, but the data from Anthropic offers valuable insights into how AI is reshaping the job market.