Less panic, more data: how will AI impact your job?
While the March jobs report was encouraging, layoffs have been making headlines as companies invest heavily in AI: just last week, Meta announced 20K cuts while Microsoft said it’s offering employee buyouts for the first time in its history.
Meanwhile, in classic Tom and Jerry fashion, OpenAI and Anthropic released reports on AI’s labor impact within days of each other.
OpenAI pushed back on the take that AI-exposed jobs are doomed. In its new jobs transition framework, which covers 900+ occupations and 99.7% of U.S. employment, the lab found that many highly AI-exposed roles might be more likely to expand or be redesigned than automated (at least in the short term).
The report states that AI exposure in and of itself is too “blunt” a measure to predict which jobs will be disrupted in the near-term. The study’s methodology expands beyond ‘technical exposure’ to include ‘human necessity’ (the strength of the need for a human worker) and ‘demand elasticity’ (measure of how much demand changes when price changes) as key variables. It validates this framework against ChatGPT usage data.
The report’s key findings:
- 18% of jobs are at a relatively higher risk for short-term automation
- 24% of jobs may see employment decline as task composition shifts while workers remain necessary for key tasks
- 12% of jobs could grow as lower effective cost could increase utilization, affordability, access, or quality-adjusted output
- 46% of jobs are likely to see less change in the short term
Taking a different tack, Anthropic surveyed 81,000 Claude users and found that while many respondents — especially those in more exposed roles — fear AI-related displacement, they also feel more productive at work.
Respondents in the highest- and lowest-paid occupations reported the largest productivity gains, most often through enhanced job scope. Interestingly, respondents who reported the greatest gains expressed more concern about job displacement. Anthropic says the survey results provide initial evidence that observed exposure, which combines theoretical LLM capability and real-world usage trends, is correlated with worker concern around AI.
And a leading indicator per Ramp data: AI’s labor substitution might start with freelance workers.
For more stories like these, subscribe to the Leading Indicators newsletter.
