By Dr. Virginia Koch
In recent years, higher education leaders at many campuses have tightly coupled student success to the personal well-being of their students. Student well-being is framed as the flourishing of the whole person, and mental health is a top concern of campus leaders. About 1 in 4 students who experience mental health challenges drop out of school, with those experiencing depression being twice as likely to leave college. This is one of the reasons why mental health risk is considered one of the top risks facing higher education, according to United Educators and Deloitte.
According to a 2024 Inside Higher Ed survey, 65% of college presidents planned to increase their institution’s capacity to meet the mental health needs of their students, faculty, and staff. However, the desire to address student mental health by using emerging technology can expose campuses to risks related to student privacy; data usage and storage, as well as ethical considerations regarding transparency; AI-usage disclosure; and informed consent.
Rapidly Shifting Approaches to Mental Health
Innovative private companies have launched technology-based solutions to address campus mental health needs with promises to increase capacity while keeping costs low. The companies provide convenient, on-demand emotional support, online counseling sessions with mental health professionals, workshops, and, most-recently, AI-driven virtual therapists or chatbots. They offer 24/7, on-demand access to virtual care and are partnering with campuses and higher education professional associations to deliver high-quality, evidence-based, inclusive care programs at an increasing rate. A 2023 nationwide college student survey conducted by a leading provider of higher ed virtual health and well-being solutions revealed that 64% of students say mental health resources provided by their institution make them more likely to graduate. That’s an enticing statistic for data-driven decision makers.
Similarly, some campus researchers are developing generative AI solutions for students interested in non-traditional counseling options. For example, Arizona State University’s Hey Sunny chatbot helps students adjust to college life, and Stanford University’s WoeBot offers conversational cognitive behavioral therapy. A 2025 groundbreaking clinical study by Dartmouth University researchers found that eight weeks of text conversations with their Therabot resulted in statistically significant reductions in students’ symptoms of depression and anxiety.
AI’s Achilles Heel
While increasingly effective, chatbots have an inherent flaw that poses a unique challenge. The flaw has been hiding in plain sight for more than 50 years. When MIT researcher Joseph Weizenbaum developed the ELIZA chatbot in the mid-1960s, he quickly observed that people interacting with ELIZA attributed human-like feelings to the computer program. Reflecting on the “ELIZA Effect” about 10 years later, Weizenbaum wrote, “I had not realized . . . that extremely short exposures to a relatively simple computer program could induce powerful delusional thinking in quite normal people.” Herein lies the risk for campus leaders deploying chatbots as low-cost alternatives to student-facing professionals and paraprofessionals.
"...the rush to adopt these technologies poses uncalculated risks for the long-term good of the humans they are designed to help."
"Do humans who use chatbots lose their right to mental privacy?"
Simply put, people crave connection. Much like the Titanic moving full-speed ahead on a dark night with limited visibility, the rush to adopt these technologies poses uncalculated risks for the long-term good of the humans they are designed to help. Instead of leaning into the desire for human connection and community, students engaging with online companions face a downward cycle that increases social isolation and loneliness but decreases the desire for human interaction. While chatbots can fill the connection gap with “artificial intimacy,” a term coined by MIT researcher Sherry Turkle, it can be counterproductive and addictive. Groomed on social media dopamine hits from “likes” and “follows,” chatbots hook users with artificial intimacy that minimizes vulnerability and offers support without challenge.
Cognitive Liberty and Mental Privacy
In addition to the dilemma of artificial intimacy, legal scholar Nita Farahany suggests unrestrained technological innovations also raise concerns about mental privacy and cognitive liberty. Every conversation with a chatbot is data. The data is stored somewhere, and it’s unclear how the data will be used. Do humans who use chatbots lose their right to mental privacy? Humans experience over 60,000 thoughts per day filled with ideas, reactions, reflections, images, memories, or ruminations. Human thoughts shape their emotions, actions, and desires. For example, if a student thinks he or she is not good at math, the belief fuels an aversion to the subject, resulting in avoidance of math-related careers.
Whether the thought is true or not, left unchallenged, their brain creates a belief system that reinforces the thought.
Higher ed leaders can address student success, mental health, and well-being with intentional, human-centered experiences. The rush to embrace AI technologies such as chatbots and virtual companions should be carefully balanced with concern for human dignity and risk mitigation. In their chapter “The Promises and Pitfalls of AI,” the authors of Encountering Artificial Intelligence wrote: “Education is not a parking lot but a journey; school ought to be thought of as a community and a place of encounter. For any education to be considered a ‘good and true’ education, it ultimately ought to cultivate a culture of encounter because ‘true knowing and true wisdom demands an encounter with reality’.”
BOTTOM LINE
Human dignity can flourish when higher ed leaders insist on robust AI data governance plans that aim to ensure data privacy, cognitive liberty, and student mental privacy, while also limiting the possibility of sensitive data breaches, inappropriate data use, and the promulgation of artificial intimacy. Strategic considerations for mental health and AI integration
Champion the preservation of the dignity of the human person by promoting mental privacy and cognitive liberty.
Ensure policies are in place to protect sensitive mental health data obtained and stored by AI-driven technology.
Convene campus experts who can consider the philosophical and moral implications of AI use, and incorporate their guidance as mental health tech solutions are implemented and monitored.
Strive for low-tech responses to human-centered issues. For example, reduce loneliness and social isolation by providing tech-free or tech-limited community spaces. Prioritize comfortable spaces that encourage in-person human interaction, collaboration, creativity, artistic expression, and care for creation.
Remember, third spaces in residence halls, academic buildings, and student unions are critical gathering places for humans to build community and engage faculty, staff, and students in intergenerational encounters.
Investing in Student Mental Health- Healthy Minds Network and ACE
United Educators Top Risk Report 2024
Significant Risks Facing Higher Education 2024 - Deloitte
A Therapeutic Relational Agent for Reducing Problematic Substance Use (Woebot): Development and Usability Study - PubMed
Leveraging AI to Support Student Mental Health and Well-Being - Higher Education Today
Survey Results - 2023 BTS
Inc. Magazine Ranks TimelyCare fastest-growing company in Fort Worth, 16th in Southwest Region - TimelyCare
First Therapy Chatbot Trial Yields Mental Health Benefits | Dartmouth
Encountering Artificial Intelligence: Theological Investigations of Artificial Intelligence (2024), Gaudet, Jerzfeld, Scherz & Wales, Eds., Pickwick Publications
Weizenbaum, Joseph, (1976). Computer Power and Human Reason: From Judgement to Calculation. W. H. Freeman. p. 7. ISBN 978-0716704645.