“Typically,” Morgan Stanley observed in a big research note earlier this week, “headcount growth has been required for revenue growth but AI is changing that relationship.” It’s the latest puzzle piece in the paradox of productivity under AI: it seems to be making work more intense, not less, and despite all the doomsday predictions of massive job loss and an impending white-collar recession, many CEOs insist they are still planning to hire more people.
The investment bank, drawing on takeaways from its annual Technology, Media & Telecom Conference in San Francisco, identified three distinct areas where AI is actually creating demand for workers—even as it threatens to hollow out others.
The findings arrive at a pivotal moment. Corporate America is increasingly signaling a “decoupling” between revenue growth and headcount growth, Morgan Stanley noted, with executives from companies like Snowflake and Shopify describing how AI tools are allowing them to do more with smaller teams. But Morgan Stanley’s analysts argue the picture is more nuanced than a straightforward displacement story—and that three areas of the labor market in particular are experiencing a surge in demand driven directly by AI.
“While some companies have reduced headcount, the majority of discussions [at the TMT conference] around AI’s impact on white collar work centered on productivity transformation and growing results without growing headcount,” they said.
In a related thought exercise last month, the Deutsche Bank Research Institute decided to ask AI how many human jobs it was going to displace, and the robot answered back that it saw 92 million jobs on the chopping block. On the other hand, it said AI would create 170 million new roles, more than offsetting the losses. “However, this transition will be disruptive,” Deutsche Bank’s Jim Reid and Adrian Cox predicted. But for Morgan Stanley, analysts said the disruption is happening right now.
Skilled Trades: The Hidden Bottleneck
The most urgent and underappreciated jobs sector, according to Morgan Stanley, is in skilled trades. The unprecedented scale of the AI infrastructure buildout—spanning data centers, power delivery systems, and networking equipment—is driving demand for electricians, electrical engineers, and construction workers that “far exceeds supply,” the bank said.
Executives from CoreWeave described a shortage of “thousands of skilled-trade workers” needed for data center construction, warning that because relevant skills take years to acquire, the supply-demand gap will persist. Nvidia CEO Jensen Huang echoed the concern, noting electrician shortages in key markets like Texas as a constraint on expansion. The bottleneck, CoreWeave noted, isn’t just about available power—it’s about having the human capital to physically deliver that power into racks and servers.
AI Training and Reskilling: A Market Exploding in Real Time
The second area of surging demand is workforce education and reskilling. As companies restructure roles around AI tools, enterprises are racing to upskill employees—and the numbers are striking. Coursera reported that AI-content enrollments reached 15 enrollments per minute in 2025, up from 8 per minute in 2024, a near-doubling in just one year.
The buyers are increasingly corporate rather than individual, with CTOs and Chief Data Officers turning to platforms like Coursera to equip their workforces with skills in generative AI, data science, and software development. Docebo, a learning management software provider, described AI as “fundamentally causing every organization to re-skill their workforce,” calling learning management systems a critical tool for delivering that training at scale.
AI Supervisors and Orchestrators: The New White-Collar Role
The third category isn’t a traditional trade or training job—it’s a newly emerging class of knowledge worker. As AI agents take over routine tasks, companies are redefining white-collar roles around supervising, orchestrating, and providing context for those systems.
C.H. Robinson told conference attendees it is being transparent with employees that “future jobs will involve managing standard operating procedures and context for AI agents rather than running operations directly.” Salesforce introduced a new productivity metric—”Agentic Work Units”—to capture the value that AI agents and the humans who manage them are delivering, as the company moves beyond measuring simple token consumption. Across industries, the message from Morgan Stanley’s conference was consistent: the workers who thrive will be those who can direct AI, not just use it.
A Tale of Two Labor Markets
The three growth areas exist alongside a more sobering dynamic for traditional white-collar employment. Snowflake cut roughly 200 positions in Q4 tied to AI-driven efficiencies, adding only a net 37 workers despite revenue reaccelerating to 30% growth. Shopify has seen headcount decline for eight to ten consecutive quarters.
Morgan Stanley frames this as a “diverging trends” story—one in which AI is simultaneously eliminating certain jobs, elevating others, and creating entirely new categories of demand that didn’t exist a few years ago. The companies and workers best positioned for what comes next, the bank suggests, are those already adapting to all three.
“Companies are increasingly allowing natural attrition to reduce staffing needs, reallocating resources toward technical talent, or shifting spending from labor to technology while maintaining headcount,” Morgan Stanley concluded. At the same time, it noted that this transition is reshaping job definitions, moving workers toward roles that supervise, orchestrate, and contextualize AI systems.
The new economy is arriving—and soon.
