Study links AI dissertation use to lower academic retention

A new working paper from the National Bureau of Economic Research (NBER) indicates that US PhD graduates in science, technology, engineering and mathematics (STEM) who used artificial intelligence to help write their dissertations are less likely to pursue careers in academia. The study, which has not been peer reviewed, suggests these graduates are instead more likely to find employment in industry.

Rising AI Use in Dissertations

The researchers analyzed dissertations from more than 226,000 STEM doctoral graduates at US universities completed between January 2019 and May 2026. Using the AI-detection service Pangram, the team assessed abstracts submitted to an online database. They categorized abstracts as either AI-written or not, without distinguishing between text that was wholly or partly generated by AI. Daniel Gross, an economist at NBER and co-author of the paper, says that 90% of the flagged AI-generated abstracts were classified by Pangram as entirely AI-generated.

The data shows a sharp increase in AI-generated text in dissertation abstracts since 2023. The presence of such text rose from 1.7% in 2023 to 18.8% in 2025. For dissertations filed during the first five months of 2026, 29.4% of abstracts contained AI-generated text. The study found that AI use was highest in mathematical and computer sciences, where 36% of abstracts had been written using AI. Additionally, about 41% of dissertations by graduates educated in non-English-speaking countries had AI-generated abstracts, compared with 23% of dissertations by those from English-speaking countries.

Impact on Career Paths

The researchers also reviewed the education and employment profiles of PhD graduates on the social-media platform LinkedIn. Among PhD graduates who completed their undergraduate degrees in the United States, those with AI-generated abstracts were 17% less likely to be employed in academia than graduates who did not use AI. Conversely, using AI made those graduates 16% more likely to be employed in industry than those who did not use it.

The authors noted that the link between AI use and landing a job in academia was less pronounced for graduates who did their undergraduate studies outside the US. Daniel Gross questions the long-term implications of this trend for scientific training. “If our students are only doing half the work and they’re dishing the rest to AI, do they really have the depth of understanding” to be good scientists? asks Gross.

Concerns About Research Training

Ang Liu, an engineer at the University of New South Wales in Sydney, Australia, highlights a specific challenge with PhD students using generative AI tools. He notes that researchers are expected to explore areas in which the answers are not already known. While AI can be helpful in areas with a substantial body of existing knowledge, Liu argues that the challenge arises when research moves into areas where knowledge is incomplete or uncertain. “That is where rigorous research training becomes even more important,” he says.

Why it matters

The findings suggest a potential shift in the career trajectories of STEM PhDs, with AI usage correlating with a move away from academic roles. As AI-generated text becomes more prevalent in scientific literature, the study raises questions about whether current training methods adequately prepare students for the independent, rigorous inquiry required in academic research. The working paper serves as an early indicator of how AI integration in higher education may reshape the scientific workforce over the coming decades.

---

Sources

• nature.com: PhD students who write dissertation with AI less likely to stay in academia | Nature

This article was drafted with AI assistance and checked against the sources above. Company claims are reported as claims. Cover image is AI-generated.