Until a couple of years ago, artificial intelligence barely factored into Keene State College’s construction safety program. Students might mess around with a few models just to familiarize themselves with AI but that was it.
Now students work with AI tools designed specifically for safety professionals, says Brady Keene, chair of the college’s safety and construction sciences department. They also complete projects meant to improve their AI competency, like giving the same task to several different models and comparing the outputs.
Keene says AI is becoming increasingly important in the occupational safety field, where it can help turn messy data into insights about how to prevent workplace injuries and deaths. And from his conversations with employers in the construction industry, it’s clear that educational institutions needs to keep up with industry expectations.
“They’ve been very upfront that if our students aren’t ready to use AI when they graduate, then they’re less attractive,” Keene says.
As artificial intelligence continues to advance, colleges and universities across NH are grappling with both new opportunities and new challenges. Many institutions are updating course offerings and curricula, experimenting with new teaching tools, using AI in research and exploring how it could improve their operations while also confronting thorny questions around ethical use, academic integrity and the technology’s impact on learning.
Whatever one’s outlook, higher education leaders say it’s here to stay and it’s critical that colleges respond. “I see it as an inevitability—something that we need to be talking about and figuring out how we are educating our students about the use of AI,” says Catherine Paden, provost and senior vice president for academic affairs at Franklin Pierce University in Rindge.
Incorporating AI into Courses
The growth of AI has prompted some schools to offer new courses, certificates or degree programs centered on the technology. The University of NH has a new AI minor as well as a certificate program geared toward working professionals who want to upskill.
Rivier University in Nashua offers bachelor’s and master’s degrees in artificial intelligence, in addition to its computer science and cybersecurity programs. The goal is to graduate students who understand the theoretical, technical and ethical dimensions of artificial intelligence, as well as how to implement it in real-world settings, says Sister Paula Marie Buley, the university’s president.
“We’re getting inquiries from different businesses: ‘How can I use AI to increase my business, to create market share, to analyze cost and price structures?’” she says. “The Rivier program is really designed for a student who wants to practice AI and apply it in different professions.”
But Buley says it’s not just the students majoring in AI who need to familiarize themselves with the technology. At Rivier, she says, students across different majors are learning to work with AI given how important the technology is becoming. “One of the primary goals of a Rivier education is to ensure students are workplace-ready upon graduation,” she says.
Donald Birx, the president of Plymouth State University and interim president of Keene State College, takes a similar view. Birx, whose 1990 disceration on neural networks mentions early forms of artificial intelligence, electrical engineering and electromagnetics, says the colleges are working toward incorporating AI into every
degree program.
That doesn’t mean it has to be a part of every class or something that every instructor uses, he says. But program leaders should make sure that, over the course of their studies, students gain enough experience to be able to use AI effectively in their field. Birx says that while most students have probably used ChatGPT or other AI programs, that’s not the same as being trained to use it effectively.
“My belief is that every student that graduates has to be competent in the application of AI to their particular areas of work,” he says.
Growing Pains
AI’s increasing adoption in the classroom has sparked concerns about its impact on academic integrity and student learning. George Siemens, the chief AI officer at Southern NH University, says the technology could cut both ways. Tools like AI tutors could improve student outcomes. But relying too much on generative AI could erode students’ critical thinking and reading skills. There’s also the temptation to cut corners or cheat. That’s why colleges need to take a thoughtful approach.
“Ultimately, the goal is learner success,” he says. “That’s the purpose we exist [for], and it’s to help learners get skilled and prepared for the work that they want to do.”
At the University of NH, many faculty members are still figuring out how to adapt assignments and assessments in the AI era, says Matthew MacManes, the vice provost for academic affairs, who chairs the university’s AI working group. Some professors have gone back to handwritten blue book tests to make sure students don’t use AI to write papers, though MacManes isn’t sure that’s the best approach for the digital age. Online courses pose another challenge. It’s much easier to make sure students are doing their own work when they’re in a physical classroom.
But MacManes says faculty members are exploring different ways to teach and assess students’ critical thinking skills. For example, oral exams that involve a discussion rather than a written answer, or exercises that allow students to use a large language model to respond to a question but then have them critique its answer.
Many institutions are also developing policies around AI use that set out expectations and guardrails. At UNH and several other colleges interviewed for this story, faculty have autonomy over whether and how to allow students to use AI in their coursework.
Some may encourage or allow it for certain assignments; MacManes says the business faculty are emphasizing AI, seeing it as a necessary skill for their graduates to be competitive. Others may not allow students to use AI at all.
What is important, MacManes says, is that the expectations are clear. “The policy doesn’t necessarily prescribe” whether an instructor allows AI use, he says. “What the policy is prescriptive about is transparency, so as to say, if a student can use ChatGPT on assignment A but not assignment B, [you’ve] got to be really clear about that.”
Dartmouth College—where researchers coined the term “artificial intelligence” back in 1956—has taken the same approach, says Jed Dobson, a professor of English at Dartmouth College and digital humanities expert who has served as a special adviser to the provost on AI.
Dobson says faculty at the Ivy League school vary in their attitudes toward AI. Some are gung-ho early adopters. Others are vocally opposed to AI for ethical, environmental or other reasons. Most fall somewhere in the middle: cautiously open to using AI, but unsure how to do so effectively and responsibly.
“Faculty need to have a lot of freedom to do what they would like to do in the classroom,” he says. “So, developing policies that respect either willingness to engage in the technology or desire to refuse is pretty important.”
Dobson wrote a report last year with a series of recommendations for AI adoption at Dartmouth, ranging from investing in more computing infrastructure to bolster AI-related research to making AI literacy a component of every degree program.
Addressing both ethical and practical concerns of AI adoption, Patrick Tompkins, president of NHTI, the community college in Concord, acknowledges AI has the potential for what has been referred to as “cognitive offloading”—the delegation of mental tasks (like data analysis, idea generation, or problem-solving) to external tools. But Tompkins says critical thinking will continue to be an essential component of education, even if it looks somewhat different.
“There’s this conversation between you and AI, and that conversation takes critical thinking,” he says. “Because you’re going to have to think ahead of time, ‘What am I trying to accomplish in this piece of writing,’ and ‘how am I going to try to accomplish that? Oh, I need to think about audience because this is for my boss as opposed to for my coworker; I need to think about persuasion because I’m trying to make an argument.’”
Enhancing Operations
Siemens with SNHU says one important question facing colleges and universities is how to ensure equitable access to AI tools. While free versions of ChatGPT and other leading models are available, the most advanced versions require paid subscriptions. Some universities, including Dartmouth, have signed agreements with leading AI firms to provide institutional access, and the Research Computing Center at UNH launched Deep Thought in 2025, a local AI platform with several free language models open to the entire university system, but that’s not a case in all places.
“There’s different populations that a university serves, different income levels, different backgrounds,” he says, “and you don’t want to have AI primarily serve one population at the expense of another.”
Birx says the overall cost of implementing AI at Keene and Plymouth would include three areas, including courses or training for faculty and staff and developing courses, integrating AI into existing curriculum, and licensing or access fees for the various models. “A specific dollar figure is not yet available because the design work is still ahead of us,” he says.
For Keene, chair of Keene State’s College’s safety and construction sciences department, the outcomes of investing in AI is not only about “ROI” viewed in terms of dollar amounts. It’s also about freeing up resources and preparing students for jobs.
“[ROI] is measured in time returned to our people. Like most public institutions today, we are operating under significant financial pressure, and artificial intelligence allows us to expand the capacity of our teams without the substantial financial burden that traditionally accompanies growth,” he says, adding the school is being deliberate about where AI is applied. “The objective is not to automate the human elements of higher education.”
Outside of the classroom, some colleges are looking at ways AI could enhance their business operations. Rivier uses it for financial modeling and has a chatbot called “Ray the Raider” that can communicate with prospective students about housing, courses and other aspects of university life.
Tompkins says he sees a lot of potential for AI to transform operations by analyzing vast amounts of student data and other information. “Why am I not, every day, getting in my inbox a solicitation from a vendor who says, ‘We can help you use AI to do X?’” he asks. “We can help you use AI to increase your enrollment by 17%. We can help you to use AI to identify students who are at risk of dropping out; use AI to analyze your budget and save a percentage.”
AI could also help colleges tailor messages to potential students who express an interest, but don’t end up enrolling. “AI can tell us, here’s who you really need to be focusing on and here’s the message you need to give them, and we’ve analyzed these data and we know this is what’s stopping that student,” he says. “And if you solve this problem for that student, they’re more likely to enroll.”
MacManes at UNH is ultimately optimistic about AI’s potential, despite all the disruption. It’s not a perfect analogy, but he says the introduction of the graphing calculator comes to mind. It automated more calculations and changed how math was taught but also unlocked new learning opportunities. “We have this new technology, which on its face seems to change everything about how we deliver to students,” he says. “But I think what’s going to happen is we’re going to see some incredible, innovative ways that we’re going to use this tool and make the way we educate students that much better. So that’s my hope.”
