BlackRock chief executive Larry Fink has warned that artificial intelligence could drive unemployment among new college graduates to its highest level in years, even without a recession, as automation disrupts entry-level office work.

Speaking at BlackRock’s Infrastructure Summit, Fink said education and training are struggling to keep pace with technological change. Recent reports have cited a 5.6% unemployment rate among recent graduates aged 22 to 27, while Handshake job postings fell more than 16% between August 2024 and August 2025, and applications per position rose 26%. Fink said expanding AI infrastructure would also create demand for electricians, plumbers, and other tradespeople. BlackRock committed $100 million to training programs intended to reach 50,000 workers over five years.

His prescription comes with a commercial interest. BlackRock led investors, including Microsoft and Nvidia, in a $40 billion acquisition of Aligned Data Centers, giving it a stake in the infrastructure requiring those workers. BlackRock is primarily an asset manager, rather than a traditional private-equity buyout firm. That’s an important variable to consider when assessing whether its investments expand employment, encourage automation, or ultimately contribute to the pressures Fink is warning against.

The danger, particularly to younger workers, is a weakened pathway to skilled employment. Junior office jobs give graduates experience that qualifications alone cannot provide. If firms automate those tasks while still demanding experienced applicants, they risk eroding the training ground for their future workforce.

Fink’s emphasis on trades reflects genuine demand, but it also aligns with BlackRock’s own infrastructure investments. Training electricians benefits workers; it also helps investors complete data centers. Those interests can coincide without making his diagnosis disinterested.

The criticism of private equity deserves its place in the conversation, as research finds substantial employment losses after buyouts of publicly listed companies, but gains after acquisitions of privately held firms. Ownership and deal structure matter; treating every investment manager as an interchangeable source of layoffs obscures the mechanisms worth examining.

The tougher question is who finances adaptation, since employers may welcome a workforce trained for their changing needs while offering fewer opportunities to gain on-the-job experience. Scholarships and retraining commitments help, but they do not settle whether companies investing in labor-saving technology will also invest sufficiently in the people expected to adjust to its deployment.

For families weighing college costs, the uncertainty changes what a degree must justify. Sure, tuition buys an education, but the financial case also depends on post-graduation job prospects. A course’s employment record and opportunities for practical experience deserve as much attention as the institution’s reputation.

Although skilled trades offer another route, demand for data centers doesn't make every training program a sound investment. Local vacancies, apprenticeship places, and the durability of construction spending matter more than a financier’s enthusiasm for the sector. Exchanging one fashionable career promise for another would leave the underlying problem intact.

Parents may also find themselves supporting graduates longer as employers become more selective, subsequently shifting some of the cost of developing workers onto households already paying for their education.

Fink is certainly right to question whether training matches demand, although companies should face an equally pressing question: how much responsibility will they accept for closing that gap?