SMART conversations about AI use: A collaborative framework for students and instructors

Generative AI presents new opportunities and challenges for psychology teachers. While AI tools can support learning, creativity, and productivity, both students and faculty have expressed concerns about their potential impact on critical thinking (Digital Education Council, 2026; Lloyd, 2025; Prothero, 2026). Although student AI use continues to rise (approaching ubiquity in some surveys), many students express confusion and uncertainty about what constitutes appropriate AI use in their academic work (Silva, 2025; Stone, 2025) and often report feeling shame and guilt over their AI use (Hamilton et al., 2025; Ling et al., 2025).

How can we help students use AI in ways that support learning rather than undermine it? Rather than focusing exclusively on detection, prohibition, or "AI-proofing" assignments, we have a responsibility as educators to help our students develop the judgment and AI literacy skills needed to make informed decisions about when, why, and how to use AI—not only for their coursework, but to prepare them for future employment.

In this article, I introduce SMART, a collaborative framework for discussing AI use that can be adapted across courses and contexts. I developed the framework as a response to feeling overwhelmed and uncertain about how to address AI in my courses. SMART organizes discussions of AI use around five key dimensions: Specific Goal, Meaningful Impact, Accountability, Rights & Risks, and Trust. The framework is intended to support reflection, foster productive dialogue, and bring conversations about AI use out of the shadows. The table below outlines each dimension, the key question it raises for students, and the AI literacy knowledge and skills students may need to engage with these questions effectively.

Psychology courses are particularly well positioned to contribute to AI literacy development because many of the issues raised by AI—including cognition, memory, cognitive offloading, bias, decision-making, ethics, and information literacy—are already directly relevant to the topics we teach. However, no single course can address all of these competencies. Developing AI literacy will require coordinated opportunities for students to build knowledge and judgment across the curriculum. SMART provides an organizational framework for thinking about AI literacy on a broad scale, while also providing instructors with a structure for navigating conversations regarding AI use for specific assignments.

SMART is intended to support conversations between instructors and students. Instructors create the conditions for responsible AI use by communicating expectations, modeling good practices, and providing guidance, while students apply judgment and take responsibility for their decisions. The sections below illustrate these complementary roles across each SMART dimension.

Specific goal 

What is the educational purpose of the assignment, and how might AI support or interfere with it? 

Instructors should clearly communicate the intended learning outcomes of the assignment and discuss which uses of AI (if any) support versus replace important learning. Students should consider whether their intended use of AI helps achieve the learning outcomes or bypasses them. While transparency has long been a best practice, discussions of assignment expectations should now also include which uses of AI are appropriate given the learning goals, course level, and student experience (see also Winkelmes, 2025; Zhou & Schofield, 2024). If AI use is prohibited, the reasons for doing so should be made clear to students, drawing on other SMART dimensions when relevant (e.g., impact on skill development or privacy risks).

Meaningful impact

How might this use of AI affect learning and skill development over time?

Instructors should help students reflect on how AI use may strengthen or weaken learning and skill development over time. Students should consider both the immediate and long-term consequences of using AI, including which skills are being strengthened, which skills may be weakened, and whether they would still be able to perform similar tasks independently in the future. These conversations can help students recognize that using AI for one assignment is not necessarily an endorsement for always using AI for that task, particularly when different assignments target different learning outcomes.

Accountability

How can students remain responsible for their decisions, outputs, and disclosure of AI use?

Instructors should clearly communicate expectations regarding acceptable AI use, disclosure practices, and academic integrity. Students remain responsible for the quality, accuracy, and integrity of their work, regardless of whether AI contributes to it. As human-AI collaboration becomes increasingly common, students need opportunities to learn how to describe, disclose, evaluate, and revise AI-generated content. If AI use is only permitted at certain stages or for specific  tasks, instructors should explain how that use should be disclosed (e.g., by sharing prompts and responses). Students should be able to explain and justify how AI was used and take ownership of the work they submit.

Rights and risks

What ethical, legal, privacy, or safety concerns should be considered?

Instructors should engage students in discussions about the ethical, legal, privacy, and professional considerations associated with AI use. Students should consider whose rights may be affected by their use of AI and whether it is appropriate to enter information into an AI system. For psychology students, these issues are particularly important because they may work with sensitive personal information, research data, case materials, or community partners during their studies and future careers. The goal is not simply to foster compliance, but to help students develop habits of responsible, ethical, and reflective AI use that will serve them as researchers, practitioners, and informed citizens.

Trust

How will students evaluate the accuracy, reliability, and fairness of AI-generated information?

Instructors should teach and model strategies for evaluating AI-generated information using credible sources and disciplinary evidence. Students should critically evaluate AI outputs, verify claims against peer-reviewed psychological literature and other credible sources, and identify inaccuracies, unsupported claims, or missing perspectives. Instructors can reinforce these skills by designing assignments that reward verification and critical evaluation rather than blind acceptance. Trust in AI-generated information should be earned through verification, not assumed because information is presented confidently.

Many questions surrounding AI use in higher education do not have a single correct answer. The educational value of AI depends not on the technology itself, but on how instructors and students choose to use it within thoughtfully designed, ethically informed learning environments (Noroozi et al., 2025). As AI becomes increasingly integrated into higher education, students and instructors will need opportunities to engage in thoughtful conversations about when, why, and how these tools should be used. SMART is intended to provide a shared framework for identifying relevant considerations, supporting productive conversations between students and instructors, and helping to organize AI literacy development across educational contexts.