Professors Are Quietly Reshaping Higher Education With AI, and We’re Just Starting to Understand How

5 min read

Picture a professor grading 120 essays at 2 AM, toggling between ChatGPT and their gradebook. Or imagine a researcher using AI to analyze decades of student feedback data in minutes rather than months. These scenarios aren’t hypothetical anymore.

New research from NPR reveals how artificial intelligence has quietly become embedded in the daily work of college professors across the United States. The investigation, based on surveys and interviews with hundreds of faculty members, shows that AI adoption in higher education is far more widespread and varied than most administrators realize. For more insights, see Education Grants and Competitions for 2026: Your Complete Guide.

According to the NPR data, approximately 40% of surveyed professors report using AI tools regularly for at least one aspect of their work. The usage breaks down into three main categories: research assistance (67% of AI-using faculty), curriculum development (52%), and grading or feedback (31%). These numbers represent a dramatic shift from just two years ago, when fewer than 10% of faculty reported any AI use in their professional duties.

Who Is Using AI, and How Often?

The research reveals significant variations across disciplines. STEM professors lead in AI adoption at 48%, followed by social sciences at 42%, and humanities at 35%. The gap is narrowing rapidly, though. English professors, initially resistant to AI writing tools, now report using them for tasks like analyzing literary patterns across large text sets or generating targeted discussion prompts from assigned readings. The pattern suggests that once faculty find a specific use case that saves real time, adoption tends to spread from there.

There’s also a generational pattern worth noting, and it’s not the one most people expect. While younger faculty (under 35) show higher adoption rates at 52%, the most interesting finding involves professors nearing retirement. Faculty over 55 who do adopt AI tend to use it more extensively and creatively than their mid-career colleagues, possibly because they feel less constrained by institutional expectations or peer judgment. They’ve already established their professional credibility, and they feel freer to experiment.

Inside the Classroom: Grading, Feedback, and Research

The investigation found that faculty members are developing sophisticated approaches that go beyond simple automation. One chemistry professor described using AI to provide initial feedback on lab reports, then reviewing and personalizing each response before it reached students. A history professor uses AI to identify common misconceptions in student essays, allowing her to address those patterns in class discussion rather than repeating the same comment across dozens of papers. These aren’t professors replacing their judgment; they’re using AI to extend their reach.

The research component is equally significant. Faculty using AI report being able to synthesize literature, identify patterns in student performance data, and generate initial frameworks for curriculum design in a fraction of the time these tasks traditionally required. For professors managing heavy teaching loads alongside active research responsibilities, those time savings translate directly into better work on both ends.

When professors use AI for grading, they’re inserting an algorithmic layer into what has traditionally been a direct human exchange. Some report feeling more objective and consistent in their assessments. Others worry they’re losing the texture of student thinking that emerges through the grading process. This tension sits at the center of how AI is reshaping the professor-student relationship in ways that aren’t yet fully visible or understood.

What Is the Psychology Behind the Secrecy?

This widespread but largely hidden adoption reveals something about how technology reshapes professional identity. Professors are experiencing what psychologists call “cognitive dissonance” between their traditional role as knowledge authorities and their new reality as AI collaborators. The secrecy surrounding much of this usage reflects deep uncertainty about what it means to be an educator in an AI-enhanced world.

Professors report feeling both relief and guilt about AI assistance. The relief comes from reduced administrative burden and enhanced research capabilities. The guilt stems from concerns about authenticity and fairness to colleagues who aren’t using these tools. Many describe their AI use as “behind the scenes” work that doesn’t affect their “real” teaching or research. This psychological separation allows them to maintain a sense of professional authenticity while adapting to new realities. What they’re actually doing is inventing new professional norms without institutional backing or peer conversation, which means those norms are inconsistent and largely invisible. For more insights, see Common Sense Education Paused Its EdTech Reviews. Here Is What Teachers Should Use Instead..

What Institutions Need to Catch Up On?

This adoption pattern illustrates how transformative technologies become normalized over time. Like the internet before it, AI is becoming invisible infrastructure rather than disruptive innovation. Professors aren’t replacing their core functions with AI; they’re augmenting them in ways that feel natural and necessary to the daily work.

The secrecy reveals how badly institutions lag behind individual adaptation. While professors quietly integrate AI into their workflows, most universities lack clear policies or support structures for this reality. This gap creates a shadow system of AI use, where individual faculty develop personal approaches without institutional guidance or peer collaboration. The result is a two-tiered arrangement where AI-savvy professors gain real advantages in research productivity and teaching efficiency, while their colleagues continue with traditional methods. That divide has implications for academic equity that universities are only beginning to recognize.

The NPR data suggests we’re past the question of whether professors should use AI. The more pressing question is how institutions will support them in doing it responsibly, transparently, and in ways that actually benefit students rather than quietly creating new forms of unequal access.

The real story here isn’t about technology replacing human expertise. It’s about humans adapting their expertise to work alongside AI, creating new forms of intellectual partnership that we’re still learning to understand. Just like that professor at 2 AM, toggling between tools to give students more useful feedback, the work is still fundamentally human. The question now is whether institutions will catch up to the people already doing it.

Related Reading

• What We Talk About When We Talk About Math
• Mandatory Media Literacy Meets the Reality of Unequal Access
• From Phone Bans to Screen Time: The Psychology Behind America’s Digital Reckoning