A new study by Anthropic, obtained exclusively by TIME, estimates that current-generation AI models could boost U.S. labor productivity growth by 1.8% per year — nearly doubling the average rate seen since 2019. Based on how people use Anthropic’s Claude AI in real-world conversations, the researchers calculated the time saved on tasks and translated that into economic value.
To arrive at this estimate, Anthropic used a tool called Clio to analyze 100,000 anonymized Claude conversations. For each task, they asked Claude to estimate how long it would have taken without AI help, worked out the wage equivalent of the time saved, and then aggregated the results across different job types to model impacts on the wider economy.
However, the study comes with important caveats. It assumes that all the hours “saved” by using AI are redirected into productive work — not into non-work activities like rest or personal time. It also does not account for the time people may spend verifying or correcting AI’s output, nor does it factor in improvements in AI capabilities beyond its current level.
Despite the optimistic productivity gains, the report sidesteps a key issue: job displacement. TIME notes that while the model doesn’t directly address unemployment, Anthropic’s own leadership has previously warned of major disruption to entry-level white-collar jobs.
