Surveys of firms show a wide spread of adoption rates, ranging from 5 percent to about 40 percent. Surveys of workers show between 20 and 40 percent of workers using AI in the workplace, with much higher rates in some occupations like computer programming.

In a preregistered online experiment, we assigned occupation-specific, incentivized writing tasks to 453 college-educated professionals and randomly exposed half of them to ChatGPT. Our results show that ChatGPT substantially raised productivity: The average time taken decreased by 40% and output quality rose by 18%. Participants assigned to use ChatGPT were more productive, efficient, and enjoyed the tasks more.

We study the effect of generative AI on productivity in a Fortune 500 firm's customer support center. Access to AI assistance increases worker productivity, as measured by issues resolved per hour, by 15% on average, with substantial heterogeneity across workers. We find that the productivity effect of AI assistance is most pronounced for workers in the lowest skill quintile, who see an increase of 0.5 in RPH, or 36%.

This study evaluates the effect of generative artificial intelligence (AI) on software developer productivity via randomized controlled trials. We report that developers with access to a generative AI coding assistant completed significantly more tasks per unit of time than those in the control group, indicating sizable productivity gains in high-skilled software work. The experimental design randomly assigned developers to AI and non-AI conditions to identify causal effects on output.

We investigate the effects of a generative AI-based conversational assistant on the productivity and quality of work of customer-support agents. Using a staggered rollout and randomized encouragement design within a large software firm, we estimate that access to the tool increases the number of resolved chats per hour by 13.8% relative to the control group. The gains are concentrated among novice and low-skill workers, with little to no effect for the most experienced agents.

This project analyzes "two field experiments with 1,974 software developers at Microsoft and Accenture to evaluate the productivity impact of Generative AI." The preliminary results "provide suggestive evidence that these developers became more productive, completing 12.92% to 21.83% more pull requests per week at Microsoft and 7.51% to 8.69% at Accenture (depending on specification)." The paper notes that prior lab experiments also found that programmers using Copilot "completed the task 55.8% faster," and cites other experiments where generative AI increased consultants’ productivity by "12% to 25%."

This BIS working paper reports on a field experiment on programmers using a large language model for coding. The authors state: "Our findings indicate that the use of gen AI increased code output by more than 50%." They quantify that "Productivity (measured by the number of lines of code produced) increased by 55% for the group using the LLM," and note that "approximately one third of this increase was directly attributable to code generated by the LLM." However, they caution that "productivity gains were statistically significant only among entry-level or junior staff, while the impact on more senior employees is less pronounced."

This working paper studies generative AI tools for sales in online retail workflows. The authors write: "We find that GenAI adoption increases sales in most workflows, with effects ranging from no detectable impact to 16.3%, depending on GenAI's role in the workflow." They describe these as "field experiments" and interpret the increases in sales as evidence that generative AI can improve sales productivity, although the magnitude varies by task and implementation.

Controlled field experiments and randomized trials document large productivity gains at the task and firm level, often alongside quality improvements. Across writing, customer support, software development, accounting, law, and translation, studies report 15% to more than 50% reductions in task-completion time, meaningful quality gains, and disproportionately large benefits for less-experienced workers. Shakked Noy and Whitney Zhang (2023) conduct a randomized experiment with 453 professionals and find that ChatGPT use reduces task-completion time by roughly 40% and increases output quality by about 18%. Erik Brynjolfsson, Danielle Li, and Lindsey Raymond (2025) examine a Fortune 500 customer-support deployment and find a 15% average productivity increase, measured as issues resolved per hour, and a 36% increase for workers in the bottom skill quintile.

Across three studies, generative AI tools increased business users' throughput by 66% when they performed realistic tasks. The article says the three studies found gains of 13.8% more customer inquiries per hour for support agents, 59% more business documents per hour for business professionals, and 126% more projects per week for programmers.

A new study on highly skilled workers found that GPT-only participants saw a 38% increase in performance compared with the control condition, and the group given GPT plus an overview saw a 42.5% increase. The article also notes that participants in the lower half of assessed skills saw a 43% increase with GPT-4 compared with baseline.

This report presents the most recent findings of Microsoft’s research initiative on AI and Productivity, which seeks to measure and understand the productivity gains associated with LLM-powered productivity tools like Microsoft Copilot. One of these is, to our knowledge, the largest, randomized controlled trial of the introduction of generative AI into organizations. Overall, the research suggests that generative AI is already aiding workers in becoming more productive in their day-to-day jobs in significant ways, including faster document creation, email handling, and analysis tasks compared with control groups.

Summarizing McKinsey’s projections, CFO Dive reports: "Generative artificial intelligence may annually boost U.S. labor productivity by 0.7% through 2040, McKinsey said in a study" and that the technology "over time will streamline at least 10% of the tasks performed by 80% of workers, and half of the tasks done by 19% of workers." The article also notes that McKinsey estimates generative AI "could increase productivity in customer care by as much as 45%," without specifying a time frame.

The proportion of employees using AI daily has risen from 10% to 12%. Frequent use, defined as using AI at work at least a few times a week, has also inched up three percentage points to 26%. The article says 38% of employees reported their organization has integrated AI technology to improve productivity, efficiency and quality.

Bloomberg reports on a field experiment by Brynjolfsson, Li and Raymond involving customer service agents for a large US company with operations in the Philippines. According to the article, access to an AI-based conversational assistant "increases the productivity of customer support agents (measured as issues resolved per hour) by 14% on average." The productivity gains were larger for less-experienced and lower-skilled workers, while more experienced agents saw smaller or negligible changes, illustrating that generative AI tools can deliver double-digit percentage improvements in real-world call-center productivity.

Despite widespread enthusiasm for generative AI, empirical evidence reveals inconsistent productivity impacts contradicting popular assumptions. AI's productivity gains are highly context-dependent, varying significantly by user skill level and task complexity. Similarly, a randomized controlled trial with 5,000+ agents at a U.S. tech support desk delivered a 35% throughput lift for bottom-quartile reps but almost no gain for veterans (Brynjolfsson et al., 2024). A 2025 meta-analysis pooling 371 estimates published between 2019 and 2024 finds no robust, publication-bias-free relationship between AI adoption and aggregate labor-market outcomes once methodological heterogeneity is controlled.

The authors estimate that AI will increase productivity and GDP by 1.5% by 2035, nearly 3% by 2055, and 3.7% by 2075. Based on studies of real-world generative AI applications, they assume labor cost savings of roughly 25 percent on average from adopting current AI tools, and note that the studies they cite find gains ranging from around 10 to 55 percent.

This industry report synthesizes data from controlled experiments and surveys on generative AI. It highlights that "The most compelling individual productivity data comes from controlled experiments" and describes a "six-month field study" where "employees using AI tools spent 25% less time on email management and administrative tasks." It notes that this 25% reduction in email processing time is "one of the most significant and measurable productivity gains documented in 2025." The report also cites surveys where "96% of employees who use generative AI feel it boosts their productivity," and references MIT research at Microsoft and Accenture showing "measurable productivity gains in software development."

This summary of McKinsey’s research explains that generative AI "could increase the impact of all artificial intelligence by 15 to 40 percent" across 63 use cases and 16 business functions. It notes that in certain functions, generative AI "could increase productivity at a value ranging from 30 to 45 percent of current function costs" and that "The marketing function demonstrates remarkable potential, with productivity increases valued between 5–15% of total marketing spending globally" while "Sales productivity could increase by approximately 3–5% of current global sales expenditures." The article also highlights that McKinsey estimates generative AI "could add $4.4T to the global economy" annually.

In this McKinsey Live webinar, partner Michael Chui explains that generative AI could add "the equivalent of 2.6 to 4.4 trillion" dollars in annual productivity growth across the 63 use cases studied. He notes that experiments on software engineers have "shown to increase the productivity of software engineers by 20, 30, 40, 50 maybe even higher percentage points," emphasizing that the technology mainly enables "partial automation of what people had previously done manually" so that their time can be repurposed to more productive work. He also states that, depending on deployment and workforce shifts, the new productivity growth rate "accelerates up to 3.3%" annually compared with historical baselines.

In a social media post promoting its research, McKinsey states that "Generative AI is poised to unleash the next wave of productivity across business." The post says AI‑driven productivity gains can have two primary effects on employees and adds that, according to McKinsey, generative AI applications in certain domains could deliver "a boost to productivity of 10% to 15%." The figure is presented as an example of potential uplift in specific business contexts rather than a guaranteed minimum effect across all employees.

Well-known workplace studies have found productivity gains that vary widely by task: some coding and writing tasks show gains above 10%, while other aggregate workplace estimates are much smaller. This means a blanket claim that adopting generative AI increases employee productivity in companies by at least 10% is not supported by a single universal estimate across all firms and tasks.