Student mental health is one of the most pressing challenges today, on college campuses everywhere. Whether AI-powered technology can be leveraged to fill the gaps in mental health care and help students thrive was the focus of a one-day event sponsored by the Center for Technology and Behavioral Health.
Held at the Class of 1982 Engineering and Computer Science Center, the AI and Student Mental Health Symposium on March 17 featured talks and panel discussions on student mental health and AI-driven innovations that can be leveraged to extend and enhance existing support systems.
In his welcome address, Andrew Campbell, the Albert Bradley 1915 Third Century Professor of Computer Science and director of the emerging technologies and data analytics core at CTBH, set up the key questions addressed at the symposium. “AI is already transforming many aspects of our lives, from how we learn to how we connect with others on campuses,” he said. “But can AI also help at-risk students thrive? Can it support mental well-being by assisting students in building resilience and agency, accessing resources, and navigating the complexities of college life?”
The talks that followed underlined the urgent need to explore innovative ways to promote well-being on campuses in the face of mounting evidence of declining mental health among young people. Panel discussions served as a platform for students to share their perspectives on mental health and experts to showcase the recent advances in AI and mobile and wearable sensing technology that can revolutionize resources that promote well-being.
Chief Health and Wellness Officer Estevan Garcia cited data from the National Healthy Minds Survey, which reported data from over 100,000 respondents from more than 170 U.S. colleges in 2023-24. While there was an overall modest decrease of 5% in both depression and anxiety scores since the previous year, challenges persist as evidenced by the numbers—38% reported symptoms of depression, 34% experienced anxiety, and 14% reported experiencing suicidal thoughts.
Despite having better scores than national comparisons in the survey, Dartmouth also has to be concerned about such challenges, Garcia said. He highlighted a number of measures aimed at countering the crisis and enhancing the well-being of the Dartmouth community.
His talk was followed by a detailed overview of the suite of counseling and wellness services available to students and staff by Matthew Duncan, assistant professor of psychiatry at Geisel School of Medicine, and Katie Lenhoff, JED campus project manager.
These efforts are part of the five-year strategic plan, Commitment to Care: Dartmouth’s Strategic Plan for Student Mental Health and Well-Being, a central piece of President Sian Leah Beilock’s vision for Dartmouth to support the mental health of graduate, professional, and undergraduate students.
Keynote speaker Ellie Pavlick, an assistant professor of computer science and linguistics at Brown University, and a research scientist at Google Deepmind, gave the audience an overview of the current state-of-the-art of large language models and their capabilities and how they could be applied to mental health.
David Blanchflower, Bruce V. Rauner 1978 Professor of Economics and an expert on happiness, presented findings from his ongoing research on the global decline in youth happiness and well-being, a shift from the previously established pattern for how most people experience happiness over the course of a lifetime.
Blanchflower first uncovered this worrying trend in the U.S., Europe, and the U.K. and has since been collaborating on a series of UN-commissioned working papers that look at happiness and well-being trends among young people around the world.
The factors causing this shift are still being debated, Blanchflower said, but understanding them is key to turning things around. His data implicates the widespread adoption of smartphones and social media about a decade ago, exacerbated by the COVID-19 pandemic. This is borne out by evidence from recent studies in Africa that in places with low internet penetration, youth report better mental health.
Blanchflower is a key organizer of a two-day symposium to be hosted by Dartmouth in October with the United Nations Development Programme’s Human Development Report Office that will bring together experts, international policymakers, and leaders in higher education to share research and best practices for addressing the decline in youth well-being.
“The bottom line, the most important question, that experts and policymakers are interested in is, what can we do?” said Blanchflower, inviting the 60 participants at the symposium to reach out with their perspectives and ideas for innovative ways to tackle the crisis.
Setting the stage for the panel discussions that followed, CTBH Director Lisa Marsch called attention to Dartmouth’s leadership in digital health, particularly in mental health.
This includes the longest-ever behavioral census study of college students using smartphones led by Campbell that won a 10-year impact award at a top conference on mobile computing and the first psychotherapy platform powered by generative AI, known as Therabot, developed by Nicholas Jacobson, associate professor of biomedical data science and psychiatry and director of the treatment development and evaluation core of CTBH at Geisel.
Daniela Armella Tangarife ’24, who is now pursuing medicine at Geisel, and Caroline Conway ’24, a clinical research specialist at Emory University, were joined by Margaret Funnell, Guarini ’97, director of undergraduate advising and research and adjunct professor of psychological and brain sciences, on a panel about student perspectives on mental health in the age of AI. Campbell moderated the panel.
The students shared experiences from their time at Dartmouth, offering the researchers in the room a window into challenges faced by students, how they learn to cope, whether support systems work for them, and how they can be made better.
Both students and Funnell agreed that the isolation during the pandemic compounded the students’ stress from academic and social pressures while learning to live independently. Finding community among peers was crucial in helping them cope. “Allowing spaces for that to happen for students is so critical and difficult,” Armella Tangarife said.
Conway believes that students are increasingly growing up in a culture of school where academic achievement is prized above self-care, and mistakes have to be minimized to ensure better future prospects.
“Students are not going to engage with resources no matter how wonderful they are, if they feel any detriment at all to their academic achievement,” she said. They worry that therapy sessions could eat into study time or that counselors may recommend taking an off-term, she said.
Funnell, who teaches a class on the neuroscience of mental illness, shared her experiences with incorporating mental health activities as part of the class, as an assignment. In their feedback students said that “this was so great, but I never would have done it if it wasn’t an assignment,” Funnell said.
The most effective measures were ones where peers were involved, according to Armella Tangarife. While the two student panelists both agreed that AI should not replace human support systems, they see value in technological innovations that can offer ways to fill the gap between demand for mental health services and resources that are currently in short supply.
The panel discussions that followed focused on leveraging new technologies to improve mental health among students. These ranged from using machine learning to predict and treat anxiety and depression, developing video games that serve as engaging digital interventions, and AI-supported solutions for improved and targeted health communication.
Panelists recognized that there are significant challenges in designing safe and effective systems that are accessible, inclusive, and effective and stressed the need for human-centric approaches.
“As we think about building interventions, I can’t overemphasize the importance of incorporating the voices, the stories, and the priorities of young people,” said Lynn Fiellin, professor of biomedical data science and a faculty member of CTBH. “The likelihood that we’re actually going to impact them, and have a positive effect, is that much greater when we co-create with them.”