New research from Anthropic estimates that current-generation AI models, such as its Claude chatbot, could significantly boost US labor productivity growth, potentially doubling the average annual growth rate seen since 2019.
The study analyzed 100,000 real conversations to calculate how AI reduces task times and its broader economic impact.
The findings highlight that AI reduces task completion time by an average of 80%, especially in sectors like healthcare assistance, where time savings can reach 90%.
This suggests that as AI diffuses across the US economy, labor productivity could see substantial gains.
What are the key findings of the Anthropic AI productivity study?
Anthropic's study found that AI currently cuts task times by about 80%, with healthcare tasks sometimes seeing as much as a 90% reduction. People using the Claude handle complex tasks that would normally take 1.4 hours and cost $55 in labor, which are now completed much faster.
The research projects that AI’s contribution could raise total factor productivity by about 1.1% annually in the US.
The researchers emphasized that this estimate is on the higher side of recent productivity projections, as total factor productivity has grown by less than 1% annually since the early 2000s.
Did you know?
Anthropic's study analyzed over 100,000 real conversations with its Claude chatbot to estimate AI's productivity gains.
How does AI reduce task completion time across industries?
AI models like Claude speed up complex, time-consuming tasks, such as drafting reports, answering service queries, and supporting healthcare workflows.
The average reduction in task time is 80%, which drastically reduces labor costs and frees up time.
In healthcare, AI assistance can save up to 90% in time by streamlining administrative and support functions.
These efficiency gains come from AI's ability to process information quickly and accurately, boosting worker speed across many knowledge-based industries.
What economic impact could widespread AI adoption have in the US?
Assuming AI reaches full diffusion across the US economy within a decade, the study projects an annual labor productivity growth increase of about 1.8%, effectively doubling rates seen since 2019.
Given labor comprises approximately 60% of productivity, this translates to a 1.1% growth in total factor productivity each year.
Such gains could accelerate overall economic growth, increase competitiveness, and reshape the labor market by changing how work is performed across sectors.
ALSO READ | Google DeepMind Hires Ex-Boston Dynamics CTO to Lead Robotics Hardware
What are the limitations and assumptions of the Anthropic study?
The study assumes all time saved by AI is redirected to productive labor, not personal activities, which may overestimate productivity gains.
It also does not factor in time spent refining or validating AI outputs, nor future AI advancements beyond current levels.
Previous studies have documented smaller improvements, such as a 14% productivity increase for customer service and 40% time savings for professional writing.
These differences highlight the challenges in accurately measuring AI’s long-term impact.
How does Anthropic's projection compare to other expert estimates?
Anthropic’s estimates are on the higher end compared to economists like MIT’s Daron Acemoglu, who predicts AI will add about 0.5% to US productivity growth annually and 1% cumulative GDP growth over the next decade.
Federal Reserve data shows generative AI adoption rising rapidly among workers, aligning with Anthropic’s projections of accelerating labor productivity growth.
Recent US data reflects 2.4% productivity growth in Q2 2025 compared to an average of 1.8% since 2019.
These findings suggest AI is becoming a major driver of US economic growth, though projections vary across research.
Looking forward, AI’s continued adoption and technological evolution will be critical in shaping the future trajectory of US labor productivity and economic expansion.
Industries and workers who adapt to leverage AI's capabilities may see the most significant benefits.
Comments (0)
Please sign in to leave a comment