AI may automate 57% of U.S. work hours by 2030, McKinsey says
NEW YORK, UNITED STATES — New research from McKinsey & Company indicates that artificial intelligence (AI) has the potential to technically automate activities accounting for 57% of current United States work hours, fundamentally reshaping the labor market.
The analysis forecasts a shift toward collaborative partnerships between humans, AI agents, and robots, with about $2.9 trillion in annual economic value at stake by 2030 if organizations successfully redesign work.
“Integrating AI will not be a simple technology rollout but a reimagining of work itself—redesigning processes, roles, skills, culture, and metrics so people, agents, and robots create more value together,” the report notes.
Vast automation potential, not mass job loss
This figure is a measure of technical potential, with 44% of hours susceptible to AI agents for nonphysical work and 13% to robots for physical tasks.
However, the firm explicitly states that this is not a forecast of job losses, as adoption will take decades and will be influenced by factors such as costs, policies, and implementation challenges.
The research emphasizes that people will remain indispensable, enabling technologies to work effectively and perform tasks machines cannot.
Historical waves of automation have seen new activities created faster than old ones were replaced, and AI is already developing new roles, such as those for refining AI agents. The outcome will depend on whether new demand and roles emerge to absorb the labor force’s shifts.
As the report notes, “Many would change as specific tasks are automated, shifting what people do rather than eliminating the work itself.”
Most human skills stay relevant but will be used differently
This study argues that AI will not render the majority of human abilities irrelevant but will fundamentally transform how they are put to use.
We have also seen that more than 70% of the skills demanded by employers today apply to both automatable and non-automatable jobs; that is, they are not irrelevant.
Since AI will perform routine jobs, employees will use their fundamental skills more often in new situations, spending fewer hours preparing and more time formulating questions and discussing findings.
McKinsey’s Skill Change Index (SCI) quantifies this shift, showing that digital and information-processing skills will be most affected, while assisting and caring skills will change the least across 7,000 skills.
High-prevalence skills like communication, problem-solving, and leadership will endure but evolve within human-AI collaborations. The most significant disruption will likely be highly specialized, automatable skills, such as specific accounting processes or coding languages.
The report states that “the overall demand for skills will depend on how the mix of jobs in the economy evolves and on how rapidly organizations adopt AI and other technologies.”
Employer demand for AI fluency surges sevenfold
Employer demand for AI fluency—the ability to use and manage AI tools—has exploded, growing nearly sevenfold in the two years through mid-2025. This growth is faster than for any other skill in U.S. job postings.
Demand for technical AI skills for building systems is also rising, though at a slower pace, with about 8 million people now in occupations that require at least one AI-related skill.
This demand, however, remains concentrated. Seventy-five percent of it is found in just three occupational groups:
- Computer and mathematical
- Management
- Business and financial operations
At the same time, the skills that AI does quite well, like routine writing and research, are being dropped in job adverts, but they are still necessary. Employers are also becoming more demanding in terms of complementary skills, such as process optimization and quality assurance, to manage AI systems.
AI agents, robots could add $2.9Tn a year
McKinsey estimates that AI-powered agents and robots could unlock approximately $2.9 trillion in U.S. economic value annually by 2030 in a midpoint adoption scenario.
Capturing this value hinges less on new technological breakthroughs and more on organizational and human adaptation.
Most possible gains are in sector-specific processes, such as clinical diagnosis in healthcare or supply chain management in manufacturing (60%).
The report reveals a significant disconnect between investment and actual benefits: almost 90% of businesses invest in AI, yet less than 40% report measurable benefits. This implies that using AI for discrete tasks within old processes is inadequate.
The greatest productivity boosts will be achieved through a total reinvention of workflows that leverage collaboration among humans, agents, and robots.
Reinventing workflows key to capturing AI’s value
The study concludes that the main obstacle to AI-powered productivity is the inability to rework fundamental work processes.
Real profits will be achieved only through reorganizing multitasking processes and enabling human-AI cooperation, rather than automating every task. This is evidenced by early adopters in sales, utilities, pharmaceuticals, and banking who use specialized AI agents to handle routine tasks, leaving humans to more valuable judgment, relationship-building, and orchestration.
This changes the way managerial work is performed, with emphasis no longer on supervising others but on coordinating hybrid systems.
Leaders are confronted with some of the most significant questions, like how to develop a culture of experimentation, safety, and trust, and how to prepare managers to work with people, agents, and robots.
Finally, investing in workers’ talent and coherent transition pathways will be as determinative as the technology itself in shaping the distribution of economic benefits from AI.
“How work evolves depends on choices made now. Investing in workers and their skills—not just in technology—will be decisive in expanding human potential and ensuring that the benefits of AI are widely shared,” the report concludes.
