Extending “GPTs Are GPTs” to Firms

We extend Eloundou et al. (2024) to build firm-level measures of exposure to large language models (LLMs) with data from two sources: Eloundou et al. (2024) for occupation-level measures of LLM exposure and Revelio Labs for firm-level employee counts by occupation. The results indicate that companies with more technology workers and AI-skilled employees tend to have higher levels of LLM exposure. We also find that differences in LLM exposure are greater between exposure categories than within them, suggesting that integrating LLMs into corporate systems may lead to significant productivity gains.