The Harvard Crimson released this week the results of a survey of the Faculty of Arts and Sciences at Harvard University. The poll shows 64% of professors say generative AI harms their courses, with 70% of respondents agreeing 'somewhat' or 'strongly' that students use the technology to reduce commitment to academic work.
The figures contrast with the student paper's own analysis: 79% of faculty in sciences and engineering, 84% in social sciences, and 63% in arts and humanities said they allow some AI use in their classes. This points to an apparent contradiction between allowing the tool and perceiving it as damaging to student development.
The survey also finds professors skeptical of student motivation. 'Roughly 70 percent of respondents somewhat or strongly agreed that Harvard students use AI,' the text reads, suggesting widespread distrust of AI as a shortcut. The study does not break down which specific uses are permitted.
For those managing tech teams, the data is telling: resistance to AI in higher education mirrors what's seen in companies, where 95% of executives report no return on investment in generative AI. The distrust isn't just about efficiency, but about the impact on building core skills.
The Week in Tech sees Harvard's survey exposing a real dilemma: AI can be a useful tool, but its use without clear standards undermines learning depth. Organizations training teams with generative AI must define where the tool supports and where it replaces human reasoning, before apparent productivity masks the erosion of technical competence.