Understanding the Value-Based Decision Making Behind Student AI Use
Authors: Yerin Seung; info@ciddl.org
In today’s classrooms, generative AI tools like ChatGPT are becoming ubiquitous. But while these tools may appear to be merely technical aids, students' decisions about how and when to use them are deeply value-based. Recent studies suggest that a complex set of psychological, academic, and social factors shapes students' use of AI. Rather than being passive users, students weigh the benefits and costs of AI use, often in the context of academic pressure, peer influence, and personal confidence. This raises important implications for educators who want to support healthy and productive use of AI in learning. This blog post synthesizes three research articles by Acosta-Enriquez and colleagues (2025), Mariñas and colleagues (2025), and Zhang and colleagues (2024) on how students conduct cost-benefit analysis to decide AI use.
Benefits and Risks of AI Use
Across the studies, students cited a range of benefits associated with AI use. Many view generative AI tools (e.g., ChatGPT, Gemini, Copilot) as powerful supports that enhance efficiency, comprehension, and academic performance. Students reported using AI to help them search for information, generate content, paraphrase text, and clarify difficult concepts, especially in STEM-related fields. These tools were seen not only as academic aids but also as sources of emotional support during periods of high stress. In one study, over 80% of students reported using ChatGPT specifically for academic assistance, suggesting that AI has become a coping mechanism for managing demanding workloads.
Yet these benefits come with significant risks. One of the most immediate concerns is the accuracy of AI-generated information. Students reported encountering hallucinated references, biased language, and outdated knowledge. In academic contexts that rely on citation and source verification, such inaccuracies can undermine the credibility of student work. More troubling is the potential for AI to erode academic integrity. Without clear norms around citation and authorship, students may inadvertently cross ethical boundaries. Perhaps most significantly, both students and researchers have raised the alarm about cognitive and psychological dependency, where excessive reliance on AI can reduce motivation, creativity, and independent thinking.
Variables Influencing AI Use and Dependency
The decision to use AI and the tendency to become overly reliant on it are shaped by a range of psychological and contextual variables. Academic stress consistently emerged as a major predictor. Students with lower academic self-efficacy—those who doubt their ability to succeed—tended to experience higher levels of stress and were more likely to use AI as a coping mechanism. In this context, AI becomes a short-term solution that alleviates pressure but may discourage deeper cognitive engagement over time.
Performance expectancy—the belief that AI will improve academic performance—also strongly influenced students’ attitudes and behaviors. The perceived ease of use (effort expectancy) played a role as well, although its direct impact on attitudes was less consistent across studies. Trust and credibility were central to students’ decisions: students who saw ChatGPT as reliable and accurate were far more likely to use it regularly. Mariñas and colleagues (2025) reported that trust was the strongest predictor of actual AI use.
Social influence added another layer. Students often learned about AI tools from peers or instructors, and the attitudes of those around them helped frame AI as either helpful or risky. In these communities, AI adoption is shaped as much by social cues as by technical functionality.
Implications for Educators and Institutions
These findings offer a clear call to action for educators and academic institutions. First, supporting students in building academic confidence and managing stress is critical. Programs that promote academic self-efficacy—through mentoring, goal-setting, and constructive feedback—can reduce the likelihood that students will turn to AI out of frustration or anxiety.
Second, AI literacy must go beyond technical skills. Students need to learn how to critically evaluate AI-generated content for accuracy, bias, and usefulness. This involves teaching them when AI can support learning, and when it might undermine it. Being able to reflect on how and why they use AI is essential for developing metacognition and academic integrity.
Third, institutions must set clear and context-specific guidelines about what constitutes appropriate AI use. Without shared expectations, students are left to navigate a gray area between support and misconduct, often under pressure to perform. By offering transparency, educators empower students to make ethical decisions.
Fourth, it’s important to engage with students’ perceptions of trust and reliability. If students believe AI is more helpful or trustworthy than human feedback, that perception should prompt reflection, not rejection. Educators should both validate students’ needs and guide them toward balanced and informed use.
Finally, schools might consider developing low-stakes tools for identifying patterns of problematic AI use. Reflections, anonymous self-assessments, or wellness check-ins can help identify when students may be relying on AI as a crutch rather than a catalyst for learning.
Final Thought
These studies remind us that students are not just passive recipients of technology—they are active decision-makers who assess risks, benefits, and values as they choose when and how to use AI. If we want students to use AI meaningfully, ethically, and effectively, we must address the psychological, social, and institutional contexts that shape their choices. The challenge is not to restrict AI use, but to create environments where students are empowered to use it wisely.
Join the conversation in our community!
CIDDL is committed to providing high-quality resources to support the increasing knowledge, adoption, and use of a range of educational technologies that can be used for educators, related services, or leadership preparation programs. For more resources, including videos and blogs, subscribe to our newsletter and follow us on YouTube, Facebook, and LinkedIn. The most important part of our CIDDL community is YOU. Join our community and share the innovative ways you are using technology, ask a question about technology integration, or participate in our bi-weekly live AI Community Chats. We look forward to seeing you in our community!