Generative AI has moved from novelty to daily tool in higher education in just a few years, and students are now using it not just for assignments, but as part of how they think about careers and skills. A new Lumina Foundation and Gallup survey of nearly 4,000 U.S. associate and bachelor’s degree students shows how deeply AI has already entered coursework and how unevenly institutions are responding.
AI use is now routine in college work
Most students say they use AI tools at least weekly for coursework, with only a small share reporting no use at all. The most common reason students turn to AI is to help break down or understand difficult material, positioning it as a learning support tool rather than just a shortcut.
At the same time, attitudes toward non-use are often tied to concerns about academic integrity, with some students avoiding AI because they see it as crossing ethical lines.
Training gaps are shaping confidence
Most students believe their schools are offering some level of AI instruction, but a significant portion say it is not enough. Perceived preparedness closely tracks with institutional policy: students at schools that discourage or prohibit AI are more likely to feel underprepared, while those at more open institutions report greater confidence.
This gap is not just academic. Students who feel undertrained are also less confident about using AI in future work environments where these tools are increasingly expected.
AI is influencing what students choose to study
The impact of AI is extending beyond coursework into major selection. Nearly half of students say AI has influenced how they think about their field of study, and a smaller but meaningful share report actually changing their major because of it.
Students in vocational and technology programs show the strongest movement, both in reconsidering their paths and in switching majors. These fields appear most directly affected by expectations that AI will change job structures and entry-level roles.
By contrast, students in healthcare and natural sciences report less impact on their academic direction.
A workforce preparation gap is emerging
The findings point to a tension between how quickly AI is entering education and how slowly institutions are adapting their approach to workforce preparation. Students are already integrating AI into daily academic work, but many are doing so without consistent guidance or clear standards.
As AI becomes more embedded in entry-level jobs and early career tasks, the lack of structured training could leave some graduates better prepared than others depending on their institution.
For higher education, the challenge is whether students will leave school knowing how to use it in a work environment where it is becoming standard.
