The AI revolution will cut nearly $1 trillion a year out of S&P 500 budgets, Morgan Stanley says—largely from agents and robots doing human jobs

The Wall Street bank estimates widescale deployment of so-called agentic AI software and embodied AI humanoid robotics could generate $920 billion in net annual benefits for companies in the S&P 500. The lion’s share of those savings, analysts say, will come from lowering payroll expenses and reducing the need for human workers in repetitive or process-heavy roles.

The projected savings equate to roughly 28% of the index’s 2026 pretax earnings—a staggering efficiency gain analysts believe will reverberate across industries. There are more caveats, as Morgan Stanley’s Thematic Investing team cautions these cost savings would “likely take many years to achieve,” and they see “significant risk” of some companies not achieving full adoption levels. The $920 billion figure represents 41% of the total S&P 500 compensation expense, they add, and they only have sufficient data to run analyses for approximately 90% of the S&P 500.

The “economic value creation,” as they put it, will come in a combination of cost cutting (e.g., lower headcount and lower costs to perform a wide variety of tasks by deploying AI) and new revenue and margin generation, as employees are freed up to spend more time on higher value-added activities that could both increase revenue and enhance margins. They see a wide variety of the balance between these two impacts, based on industry and occupation. The $920 billion in annual economic benefit could translate into a $13-$16 trillion boost in market value for the S&P 500, according to the report, depending on valuation multiples. That figure would amount to nearly a quarter of today’s entire market capitalization.

Sectors most exposed

Not all industries will feel the effects equally. As you can see from the chart below, Consumer staples distribution and retail, real estate management, and transportation are among the most exposed sectors, with potential AI-driven productivity benefits exceeding 100% of forecast 2026 earnings. Healthcare equipment and services, autos, and professional services also face major disruption and opportunity.

By contrast, industries that already run lean on labor relative to earnings—such as semiconductors and hardware—show comparatively lower AI value potential.

Jobs at risk, new roles ahead

Though the topline savings will come from payroll reductions, Morgan Stanley stressed the distinction between full automation and task-level augmentation. Agentic AI, which spans generative AI and software applications, tends to reassign tasks rather than eliminate jobs outright, while embodied AI in the form of humanoid robots poses more direct substitution risks in industries such as logistics and physical retail.

The report also anticipates entirely new categories of jobs—from chief AI officers to AI governance specialists—emerging alongside the displacement trend, echoing earlier waves of technological disruption that boosted demand for programmers, IT professionals, and digital marketers.

A long ramp-up

Despite the headline number, the analysts caution full adoption is likely to unfold over years, if not decades. Firms will lean first on attrition and process efficiencies rather than immediate mass layoffs, particularly in sectors where customer-facing roles drive revenue.

Still, the message for investors is clear: AI is no longer a speculative theme. The cost savings potential is so large it could become one of the most powerful drivers of U.S. corporate earnings growth in the second half of this decade.

For this story, Fortune used generative AI to help with an initial draft. An editor verified the accuracy of the information before publishing. 

Introducing the 2025 Fortune Global 500, the definitive ranking of the biggest companies in the world. Explore this year’s list.