OpenAI has launched GDPval, a new benchmark designed to measure the economic value generated by artificial intelligence models across various industries and professions. This initiative, introduced on September 25, 2025, aims to assess AI's practical impact on real-world economic output, moving beyond theoretical capabilities.
The GDPval benchmark evaluates AI performance in nine key sectors that contribute to the U.S. Gross Domestic Product, covering 44 distinct professions. In its initial assessment, subject matter experts compared AI-generated reports with their own work. The results indicated that OpenAI's GPT-5 model was rated as superior or on par with human experts in 40.6% of the evaluated cases. This marks a significant improvement from GPT-4o, which achieved a 13.7% success rate approximately 15 months prior.
While Anthropic's Claude Opus 4.1 slightly outperformed GPT-5, achieving a "not inferior" rating in 49% of tasks, OpenAI noted that Claude's higher score may be partly attributed to its generation of more visually appealing charts rather than solely superior performance in core tasks. The GDPval benchmark seeks to provide a more grounded understanding of AI's capabilities, focusing on tasks with tangible economic implications.
This aligns with sentiments from industry leaders like Microsoft CEO Satya Nadella, who emphasizes that AI's ultimate success should be measured by its contribution to real-world GDP growth and firm expansion, rather than solely by benchmark scores. Historical technological shifts, such as the adoption of electricity or computers, have shown that widespread economic integration requires considerable time and societal adaptations.
OpenAI acknowledges that the current GDPval test, which focuses on report generation, is an initial step. Future iterations are planned to incorporate a wider array of industries and more interactive workflows to provide a more comprehensive evaluation of AI's potential across diverse professional landscapes. The insights from GDPval are expected to inform businesses, policymakers, and consumers about the evolving economic contributions of AI, fostering a clearer understanding of its transformative power across sectors like healthcare, finance, and manufacturing.