A global survey of 415 CEOs shows how quickly AI is changing the bottom of the corporate ladder. Entry-level work is not vanishing overnight, but the training pipeline that creates future senior talent is already being narrowed.
The new signal from the C-suite is blunt. According to a report from the Oliver Wyman Forum and the New York Stock Exchange, 43 percent of CEOs plan to shift away from junior roles over the next one to two years, up from 17 percent in 2025. Thirty-three percent are shifting toward midlevel roles, while 45 percent expect to hold headcount flat and 29 percent plan cuts of more than 5 percent. In technology, media and telecom, the planned reductions are especially sharp, with 43 percent of CEOs expecting cuts.
That does not mean every company sees junior talent as disposable. Oliver Wyman found that AI ROI leaders are more likely than laggards to increase junior hiring, with 24 percent shifting toward junior workers. Their logic is that AI can make new hires productive sooner, not make them unnecessary. But that group is still the exception. The larger pattern is a corporate workforce becoming more experienced, more middle-heavy and less open at the base.
The apprenticeship pipeline is at risk
The most important risk is structural. Companies still need midlevel employees who understand customers, workflows, internal politics and judgment calls. Those people do not appear fully formed. They usually start in junior roles, learn the company, absorb mistakes at lower stakes and move upward over time.
Recent academic research makes the problem harder to dismiss as speculation. A Harvard-linked working paper analyzing resume and job-posting data covering roughly 62 million workers across 285,000 US firms found that junior employment at companies adopting generative AI fell by about 7.7 percent after six quarters relative to non-adopting firms. Senior employment, by contrast, did not fall in the same way.
The mechanism matters. The decline was driven mainly by reduced hiring, not mass layoffs. Companies are not simply firing large numbers of junior staff. They are bringing in fewer people at the bottom. That is quieter, but it may be more damaging over time because it erodes the path into professional work before young workers ever get a chance to prove themselves.
Insurance shows the pressure in real time
The insurance sector offers a useful case study because its talent problem is already visible. The Q1 2026 Insurance Labor Market Study from The Jacobson Group and Aon found that job openings in finance and insurance fell to their lowest monthly level in a decade by December 2025, dropping from an annual average of 281,000 openings to roughly 138,000 in a single month. Automation requiring fewer staff was the most common reason cited by firms reducing headcount.
At the same time, the industry faces a retirement wave. Long-running projections have warned that hundreds of thousands of insurance workers could leave the sector by 2026, creating pressure on carriers, brokers and agencies to replace institutional knowledge. That is where the AI story becomes complicated. The administrative and junior analytical jobs most exposed to automation have also been the training ground for underwriters, claims specialists and account managers.
If those roles shrink too quickly, the industry may solve one cost problem and create a bigger capability problem. Insurance depends on judgment built through repetition: reading policies, spotting exceptions, understanding risk, and knowing when a clean-looking case is not clean at all. AI can speed up parts of that work, but it does not automatically replace the human learning process that turns a junior employee into a trusted decision-maker.
The succession planning contradiction
The CFO data points to the same tension. In a companion Oliver Wyman Forum and NYSE survey, 70 percent of CFOs said they plan to intensify succession planning and leadership development over the next three years. Yet 64 percent also expect the finance function to shift away from junior roles, and only 13 percent expect growth in junior hiring.
Those two goals are difficult to reconcile. A company cannot build a deep leadership bench while cutting off the entry points that feed it. The traditional finance pyramid, wide at the base and narrow at the top, is being pushed toward a middle-heavy structure. That may look efficient in a spreadsheet, especially while AI handles routine reporting, reconciliation and analysis. But it leaves a practical question unanswered: where will the experienced finance leaders of the 2030s come from?
What this means for startups
For startups, the lesson is not simply that hiring will get harder. It is that the definition of junior talent is changing. Founders can no longer assume that larger companies will train early-career workers for several years and then release them into the startup labor market. If big employers hire fewer juniors now, the future pool of experienced operators, analysts and engineers will be thinner.
There is also an opening here. AI-native juniors who can use tools well, question outputs and take on analytical work earlier may become more valuable, not less. As Mark Cuban told Business Insider, "What people thought of as an entry-level position before, show up and do the tedious work, is gone." His point was not that young workers have no role. It was that the old bargain has changed. Companies now expect new hires to arrive ready to contribute in a workplace where the easy tasks are increasingly automated.
The next few years will test whether companies can rebuild apprenticeship around AI-augmented work, or whether they will weaken their own future talent supply in pursuit of short-term efficiency. The CEOs surveyed have made their direction clear. The consequences will show up later, when today's missing junior hires should have become tomorrow's managers.
