The cost of earning a four-year degree has never been steeper. In-state students at public universities now spend roughly $30,000 per year on tuition, fees, housing, and other expenses, putting the four-year total near $120,000. At private nonprofit colleges, that figure climbs to around $252,000 over four years. Factor in student loan interest over a typical 20-year repayment window, and the true financial burden grows substantially. For people building skills in artificial intelligence (AI), however, a traditional degree may be less essential than it once was.
PwC’s 2025 Global AI Jobs Barometer, released in June 2025 and based on analysis of close to a billion job ads from six continents, offers a striking finding on credentials: “Employer demand for formal degrees is declining particularly quickly for jobs exposed to AI, especially jobs more highly automated by AI.” Between 2019 and 2024, the share of AI-augmented job postings that require a degree fell 7 percentage points, dropping from 66% to 59%. For roles where AI automates tasks directly, the decline was steeper still, falling 9 percentage points from 53% to 44%.
The same report identifies the sectors where AI-skilled workers are best positioned: Information and Communications Technology, Professional Services, and Financial Services lead the way, a finding drawn from AI job postings tracked between 2012 and 2024. Employers in those fields are not simply hiring more people; they are paying a premium for the right skills. Workers with advanced AI capabilities commanded a 56% wage premium over peers in the same roles who lacked those skills, up sharply from a 25% premium the year before.
The report also documents how quickly the nature of AI-exposed work is shifting. The skills sought by employers in occupations most exposed to AI are changing 66% faster than in other occupations, an acceleration from 25% the prior year. That pace creates a real challenge for anyone whose training is more than a few years old. PwC notes that 100% of industries are expanding their use of AI, including sectors not commonly associated with the technology, such as mining and construction. Workers who are not actively updating their skills face an increasingly narrow path in roles where AI is already embedded in the workflow.
The PwC findings are consistent with earlier research from Goldman Sachs, which analyzed the task content of over 900 occupations and estimated that roughly two-thirds of U.S. jobs are exposed to some degree of AI automation. Of those exposed occupations, Goldman Sachs economists estimated that between a quarter and half of their workload could ultimately be replaced by AI. The firm also projected that generative AI could expose the equivalent of 300 million full-time jobs globally to automation over time.
The two bodies of research differ in scope, sample size, and methodology, but they converge on a central point: AI will reshape the workforce in ways that are already measurable. The PwC data offers one note of reassurance on the near term. Jobs in virtually every category of AI-exposed occupation are still growing. Between 2019 and 2024, occupations with lower AI exposure saw strong growth of 65%, but growth remained robust even in highly exposed occupations, at 38%. AI is reorganizing work rather than eliminating it outright, at least for now.
What remains genuinely uncertain is the trajectory over the next decade. Productivity growth in the industries most exposed to AI has nearly quadrupled since 2022, rising from 7% to 27%, while productivity in industries least exposed declined slightly. That divergence could widen the gap between workers who adapt and those who do not. Whether AI ultimately drives broad unemployment or simply makes the workforce far more productive in aggregate is a question neither study fully resolves.
These States Have the Highest Student Loan Debt per Capita
∴
Editor’s note: This article updates the four-year college cost estimates to reflect 2024-25 College Board data, corrects the characterization of PwC’s 66% figure (which measures the pace of skills change in AI-exposed roles, not job growth), and adds new PwC findings including the 56% wage premium for AI-skilled workers, specific degree-requirement decline figures (7 and 9 percentage points), and the productivity growth comparison between AI-exposed and AI-insulated industries. The Goldman Sachs 300 million jobs estimate and the “over 900 occupations” scope of that analysis were also clarified.
Contact [email protected] for any questions or corrections.