Morgan Stanley says AI will require preparation for jobs that don’t even exist yet

Tech giants and stock market investors are increasingly aligned in predicting that artificial intelligence will permanently eliminate millions of office jobs and make traditional work obsolete.

Software and services stocks have been falling sharply, with software multiples down about 33% since the end of 2025 as investors worry about AI’s potential to automate vast swaths of intellectual work. Earlier this year, Elon Musk predicted that AI and humanoid robots will make work completely “optional” within the next 10 to 20 years, ushering in a post-scarcity economy in which money itself becomes irrelevant.

Also read: European Central Bank still not seeing wave of AI-driven layoffs

He joins a growing chorus of technology executives issuing blunt warnings about human obsolescence; OpenAI CEO Sam Altman recently warned that superintelligence could soon surpass even top corporate executives, while Microsoft AI chief Mustafa Suleyman and Anthropic CEO Dario Amodei projected that widespread automation of office jobs could arrive in one to five years.

Economists remain skeptical about the time frame, noting that the apocalyptic narrative could be as much a tool to justify astronomical valuations for technology companies as it is an imminent economic reality.

But a new cross-sectional research report on different asset classes from Morgan Stanley offers a surprisingly down-to-earth message for anxious employees and nervous markets: Most of you won’t be permanently unemployed; you’ll just find new jobs, many — or most — of which don’t even exist yet.

In addressing the widespread concern that AI will “replace millions of jobs and increase unemployment by an equivalent amount,” a large team of analysts at Morgan Stanley pointed directly to history.

Over the past 150 years, profound technological changes — from electrification and the tractor to the computer and the internet — have fundamentally transformed the workforce, but “have not replaced human labor.”

When the spreadsheet became popular in the 1980s, for example, it automated repetitive financial modeling tasks and reduced the need for certain accounting assistants.

At the same time, however, it freed up analysts’ time for more complex activities and gave rise to entirely new financial professions. Likewise, the bank argues, AI will only change “job types, occupations and required skills.”

“Although some functions can be automated, others will be enhanced with the complementary use of AI, and completely new ones will be created,” the report states. Rather than a mass extinction event for office workers, the bank sees the corporate environment as just gearing up for an evolution.

What jobs will come?

So what will these new jobs look like? Morgan Stanley outlines several emerging professions that it predicts will soon become commonplace in companies.

As AI becomes central to business strategy, companies are expected to hire high-level executives, such as chief AI officers, to guide adoption of the technology across departments.

There will also be a strong increase in AI governance roles focused on data compliance, policy oversight and information security, especially in sensitive sectors like healthcare.

The technology sector may see the emergence of hybrid roles, such as the combination of product manager and engineer. Driven by natural language programming tools, product managers will increasingly practice so-called “vibe coding” — creating prototypes and refining concepts on their own before handing them off to engineers for implementation.

Highly specialized roles can also arise in different sectors. In the consumer segment, “AI-powered personalization strategists” and “AI-powered supply chain analysts” will combine data science and customer experience.

In industry, we will see “predictive maintenance engineers” and “smart power grid analysts”, while healthcare will demand “computational geneticists” and specialists dedicated to overseeing AI diagnostics.

For financial markets, the current panic over AI disruption appears premature, if not outright misguided, in the bank’s assessment. Morgan Stanley notes that services sectors and cyclical segments that have recently performed well below average due to disruption fears only account for about 13% of the S&P 500’s market value.

Fortune has previously reported a similar finding from other Wall Street economists: the market appears to be panicking on its own without fundamentals justifying it, a trend likely worsened by the growing number of retail investors in the stock market.

Apollo Global Management chief economist Torsten Slok warned that “the entire market (is) exposed to a large move,” arguing that the share of S&P 500 companies that move more than 10% in a single day has increased, while options activity remains “extremely elevated, consistent with heavy speculation from retail investors and similar exposure to leverage.”

This makes the market structure “more fragile and more vulnerable to an abrupt and disproportionate movement”.

But what if this time is different?

Morgan Stanley’s report offers welcome relief — but it may be telling a comforting story that doesn’t fit the technological and economic realities of 2026.

While it is true that previous waves of automation have created as many jobs as they have destroyed, AI may represent a qualitatively different shift, by targeting cognitive, creative, and decision-making tasks previously considered immune to automation.

In a new paper released on the same day, two Nobel Prize-winning economists (Daron Acemoglu and Simon Johnson) and another hugely influential one (David Autor, known for his work on the “China Shock”) argued that this time might actually be different.

In “Building Pro-Worker Artificial Intelligence,” published by the Hamilton Project, they warned that “pure automation technologies” do the opposite of collaborating with workers: “They commoditize human expertise, making it less valuable and potentially superfluous.” The specific stock of specialized human knowledge may become “obsolete” with the widespread adoption of this technology.

Although Morgan Stanley’s thesis reflects historical optimism, the lessons of the past may not directly apply in a context of transition from tools that augment human work to systems that replace cognition.

As warned in Citrini Research’s speculative essay, AI could generate productivity gains that further decouple corporate profits from employment than occurred in the computing era. If companies can expand production with largely automated workforces, they will have little incentive to rehire at historic levels.

Morgan Stanley cites evidence that American companies are already reaping concrete benefits from adopting AI. In the fourth quarter of 2025, 30% of companies identified as AI “adopters” reported quantifiable financial or productivity benefits from the technology, up from just 16% a year earlier.

As a result, expectations for future profit margins are actively accelerating at companies successfully utilizing AI. The continued increase in these margins, and the number of new jobs created by these companies as a result, will indicate whether Morgan Stanley’s forecast is correct.