AI Reshapes Computer Science Job Market: Graduates See 13% Earnings Drop Amid Shifting Demand

Computer Science Graduates Hit by AI-Driven Market Shift

Computer science graduates are facing their toughest job market in years, with employment and earnings prospects deteriorating at levels comparable to those who graduated during the Great Recession. According to a working paper by three U.S. Census Bureau economists, the most AI-exposed fields have seen initial employment likelihood decline by five percentage points, while full-quarter initial earnings dropped by 13 percent since ChatGPT launched in late 20221

1

. The generative AI boom has fundamentally altered job prospects for computer science graduates, with many finding themselves competing in a transformed landscape where traditional entry-level positions are becoming scarcer.

The impact extends beyond computer science to mathematics, statistics, engineering disciplines, networking, and journalism—all classified as AI-exposed fields. Unlike typical recessions that spread effects evenly across industries, this shift disproportionately affects specific majors while leaving less AI-exposed fields like nursing and education largely unaffected1

1

. The Census Bureau research reveals that about half of the earnings decline stems from graduates shifting into lower-paying sectors like retail and food service as traditional tech opportunities contract.

Business Uncertainty Reshapes Hiring Patterns

The job market challenges facing computer science graduates extend beyond direct AI replacement. Firms may be hiring fewer recent graduates due to business uncertainty surrounding AI's future impact, even when they aren't actively deploying the technology themselves. The Census Bureau economists noted that employers concerned about AI making it harder to find suitable candidates or uncertain about future business models are pulling back on entry-level hiring1

1

.

Enrollment patterns suggest the market is self-correcting to some degree. The most AI-exposed majors experienced rapid growth in graduates leading up to ChatGPT's release, but those numbers have started leveling off quickly afterward. Survey data on STEM departments indicates enrollments in some of the most exposed programs have begun declining1

1

. However, the number of graduates finding jobs in their field has fallen even faster than enrollment adjustments, highlighting the depth of the market transformation.

Underemployment Remains Lower Despite Headlines

While computer science graduates have experienced unemployment rates a couple of points higher than average since the launch of ChatGPT, they continue to face substantially lower underemployment than most peers. This critical distinction means CS graduates are still more likely to land professional, college-level roles that meaningfully use their computer science degrees, including well-paying positions across the tech sector2

2

.

Recent research from the National Association of Colleges and Employers confirms that computer science still tops the list of highest projected salaries for new graduates and ranks third in overall employer demand. A national survey conducted by Strada found that employers rated math-intensive STEM graduates, including those with computer science degrees, as the most difficult to attract for entry-level positions and the highest for critical thinking across all fields2

2

. These metrics suggest the field is transforming rather than collapsing.

Forward-Deployed Engineers Emerge as New Opportunity

The rise of AI tools has created demand for a new category of talent: forward-deployed engineers who can sit inside organizations, apply AI tools to specific business problems, and rebuild workflows around the results. As AI becomes more sophisticated and accessible, a version of the Jevons paradox is playing out—when a capability becomes cheaper and more efficient, demand for that capability expands rather than contracts2

2

.

Multibillion-dollar investments across venture capital and private equity are betting on this premise—that the next wave of value creation will be built by people who know how to direct AI effectively. Computer science graduates remain among the best positioned to fill these emerging roles, though the talent pipeline to meet this demand does not yet exist2

2

. The computer science jobs of tomorrow will rely less on fluency with easily commoditized tools and models, and more on the ability to leverage AI to fundamentally rewire how organizations operate.

Long-Term Implications Remain Uncertain

The Census Bureau researchers acknowledge concern about the magnitude of their findings but hesitate to extrapolate short-term trends into long-term labor market forecasts. Historical evidence shows many jobs performed today didn't exist decades ago, making it difficult to predict future roles or required skills. Previous research on those who entered the workforce around the Great Recession suggests years of poor job prospects can leave graduates behind the curve on earnings, though the AI-driven transformation may create entirely new career paths that current data cannot capture1

1

.

Changes in self-employment and graduate school enrollment could account for as much as half of the five-percentage-point employment decline, suggesting AI is one factor among several reshaping early career outcomes. The researchers emphasize caution in speculating about AI's economy-wide impact based on shorter-term effects on a subset of workers, even as the data confirms significant near-term challenges for computer science graduates entering today's job market.