<p>The accelerating integration of artificial intelligence (AI) into higher education offers clear opportunities for learning support while also raising concerns that bear directly on student well-being; in particular, although AI-enabled tools such as ChatGPT can widen access to information and on-demand assistance, their implications for students’ emotional well-being, especially their capacity to tolerate uncertainty and regulate emotions, have yet to be examined in sufficient depth. Guided by the concepts of Situated Learning Theory, the present study investigated the possibilities of a structured ChatGPT-assisted instructional routine through an individually randomized experimental pretest-posttest control-group design, drawing on a convenience sample of 60 undergraduate students from a state university. Participants were randomly allocated to an experimental group (<i>n</i> = 30), which engaged in an instructor-orchestrated ChatGPT-assisted learning routine over a 12-week semester, or to a control group (<i>n</i> = 30), which received traditional instruction. Data were collected via pre- and post-tests assessing intolerance of uncertainty (IU) and emotion regulation (ER). Independent samples <i>t</i>-tests were used to examine pretest group comparability, and ANCOVAs controlling for pretest scores were used as the primary analyses of posttest group differences. The baseline-adjusted results indicated lower posttest IU and higher posttest academic ER among students in the experimental condition. These findings indicate that participation in a carefully structured ChatGPT-assisted instructional package can support cognitive growth and strengthen students’ emotional resilience, with potential to ease anxieties that often accompany higher education and to encourage the deliberate integration of AI tools in ways that foster more supportive and emotionally intelligent learning environments.</p>