The Texas A&M University System is launching a targeted AI intervention to support Pell-eligible first-year students across six diverse campuses. The initiative aims to close completion gaps by using behavioral science and real-time guidance to help low-income and first-generation learners navigate complex financial aid and degree requirements. System leaders have set a mandate to scale system-wide by 2030 and raise persistence rates by 2.5 percent for Pell-eligible and first-generation students in the system.
Leaders at the participating institutions are deploying chatbots to provide automated, 24/7 guidance to bridge the gap when human staff are unavailable. The strategy offloads high-volume "transactional" hurdles — such as registration challenges and financial aid complexities — allowing university staff to focus their capacity on high-touch, “transformational human interventions” and crisis support, said Texas A&M University System Vice Chancellor of Academic Affairs and Chief Transformation Officer Shonda Gibson.
The rollout serves as a deep "institutional study" in student success, forcing a rigorous audit of campus-specific infrastructure. By using real-time student feedback as a translator between traditional academic bureaucracy and the modern learner, the System aims to identify and correct localized roadblocks to ensure the university environment is accessible and culturally grounded for all students.
Texas A&M University System leaders emphasize the importance of the "human at the center" of all broad-scale technology solutions.
The bigger picture:
More and more institutions are exploring ways technology, including artificial intelligence, can support smarter, more streamlined operations on campus by allowing staff to focus on the human impact versus answering transactional questions. What is unique about the Texas A&M University System approach is that it forms a cohort of six vastly different institutions — from a flagship to a large, public historically Black university, to smaller regional campuses — whose leaders meet weekly to share insights and information about what they’re learning through students’ engagement with the bots.
“If you’re all alone doing hard work, that’s one thing, but if you’re part of a [group]...and you can compare notes and share things,” institutions have an opportunity to approach student success from a systematic level, Gibson said.
So far, the data shows students are talking to the chatbots the most between the hours of 11 p.m. and 3 a.m. — times that no staff members would have been available to meet with them. They’re helping students with everything from helping them to register for classes, referring them to other schools within the system or through their course exchange program when the course isn’t available at their local campus, to referring them to the Regents’ emergency grant program when they’re worried about expenses that might cause them to stop out, to mental health triaging.
It’s all about asking, “How can we best serve a vulnerable student population that may not come to a human and ask these questions and really focusing on our own internal transparencies – what are we learning from the students that we have roadblocks and barriers and bumps to student success,” she said.
The participating institutions had to apply and demonstrate a level of institutional readiness to be accepted into the program, which is still built on the idea of student success being driven by humans at the center.
“You wouldn’t want to launch it in a place where you just didn’t have the human capacity to do so. You had to show that you had a team that you could put together to really focus on this,” Gibson said, adding that leadership teams at each campus had to not only be involved in programming the bot’s language to reflect the culture of each campus, but also audit the university’s infrastructure around how and where information is stored in order to help the bot provide smart answers to students. “It’s not one-size-fits-all. It’s a lot of work.”
The other unique thing about TAMUS’ engagement with AI is that the system they are using, which is powered by Mainstay, is the same technology being used by the Texas Higher Education Coordinating Board to provide advising to high school seniors and adult learners across the state, providing a more coordinated experience across the sectors. Gibson said this allows them to see where students are coming in from high school and provide a continual experience that aligns with their university studies.
One thing Gibson said hasn’t come up on the Texas campuses yet is how the bot engages with students who might report sensitive concerns around political hot buttons, like immigration status, but this is one area campus leaders should consider before launching similar wide-scale solutions. If the goal is to ensure the “safety and security of the engagement" with students and the “humans at the center” on campus, leaders should work to understand how the technology would potentially engage with government agencies as reporters — especially when deploying solutions at a state-wide level. Otherwise, the same level of trust leaders rely on to inform them of financial or family emergencies that are impacting students’ ability to persist could backfire in states where governors prioritize cooperating with the federal government’s deportation agenda.
