AI Becomes Common in Schools While Confusion Grows Over Cheating and Integrity

AI use in U.S. schools is skyrocketing among students and teachers, but clear training, policies, and guidance lag far behind. The RAND report warns that without urgent action, confusion over cheating, critical thinking, and responsible use could erode trust and learning.

The RAND Corporation, working in collaboration with the Center on Reinventing Public Education and the MIT Teaching Systems Lab, has unveiled a report that captures the whirlwind of artificial intelligence sweeping through American classrooms. Based on eight nationally representative surveys and a series of administrator interviews, the study depicts a system where AI is becoming central to teaching and learning but remains poorly governed. The findings highlight a striking contrast: while students and teachers are embracing AI tools at an unprecedented pace, schools and districts are still struggling to define boundaries, establish rules, or provide proper training.

Students and Teachers Dive In

By early 2025, 54 percent of middle and high school students reported using AI for schoolwork, a sharp rise of more than 15 percentage points in just a year. High school students were leading the surge, with 61 percent admitting they relied on AI, compared to 41 percent of middle schoolers. Teachers have moved just as quickly. More than half of English, math, and science teachers reported using AI for lesson planning or teaching, up dramatically from just one in four the year before. This explosion of use signals not only experimentation but growing dependence on the technology to handle both learning and administrative tasks. Yet as AI becomes normalized, students are left largely to their own devices in determining what counts as responsible use.

Training Trails Far Behind

The growth in AI use has not been matched by growth in guidance. Only 35 percent of district leaders said their schools had provided students with any training in AI use, and just 19 percent of students recalled receiving instruction from teachers. The overwhelming majority, more than four out of five, said they were figuring it out on their own. Teachers themselves were little better supported: about half reported receiving some professional development on AI, but many said it was too superficial to be useful. The absence of systematic training has created a patchwork of practice. In some classrooms, AI is treated as a powerful learning tool, in others as a threat, and in many as something students use informally without clear expectations.

Anxiety and Ambiguity

The lack of clarity has fostered confusion and unease. Parents and students are far more skeptical than administrators about AI’s impact on learning. Sixty-one percent of parents worried that greater reliance on AI would erode students’ critical thinking skills, a fear shared by 55 percent of high school students and nearly half of middle schoolers. District leaders, however, were far more optimistic, with only 22 percent sharing the concern. Many leaders described AI as an essential skill for future employment and higher education, likening it to literacy in digital tools. Some expressed excitement about its ability to enhance creativity and streamline classroom work. For students and parents, however, the line between using AI to learn and using it as a shortcut remains dangerously blurred.

Nowhere is that ambiguity sharper than in the debate over cheating. Asked whether using AI for schoolwork is dishonest, 77 percent of parents said “it depends,” reflecting a deep lack of consensus. Less than half of principals reported having any AI use policy in place, and only a third of teachers said their schools had rules about AI and academic integrity. Where policies exist, they are often vague or limited, leaving students in a fog of suspicion. About half reported being worried they might be falsely accused of cheating with AI, and one in six said they either had been accused or knew someone who had. The murky rules are fueling tension in classrooms just as AI becomes ubiquitous.

What Schools Must Do Next

RAND’s researchers stress that schools cannot continue in this gray zone. They call on states and trusted education agencies to take the lead in drafting clearer frameworks that districts can adapt locally. Guidance must make explicit the difference between using AI as a crutch, which undermines learning, and using it as a complement to critical thinking and creativity. Concrete examples of acceptable and unacceptable uses, they argue, could reduce anxiety and restore trust.

The report also warns against overlooking elementary schools, where AI training is almost nonexistent. Only 3 percent of district leaders reported offering AI instruction at this level, even though nearly half of elementary teachers already experiment with AI tools. RAND argues that introducing responsible AI use early is crucial, both for building digital literacy and for shaping the habits that will carry into middle and high school. Without early foundations, the challenges now seen in high schools, confusion, suspicion, and uneven use, are likely to intensify.

A System at a Crossroads

Interviews with district leaders reveal administrators who recognize the urgency but remain unsure of the path forward. Some districts have begun rolling out training programs, ranging from introductory “AI 101” sessions to deeper dives into classroom applications. Others are drafting policies that give teachers discretion over what counts as cheating, while several leaders openly called for stronger direction from state education departments. Across these responses is a shared recognition that AI is here to stay and that ignoring it is no longer an option.

Ultimately, the RAND report presents a paradox. In just two years, AI has gone from novelty to normal in American classrooms, but the policies, training, and cultural frameworks to support its responsible use have lagged badly behind. Students, parents, and teachers are left uncertain about what is allowed, what is dishonest, and what this revolution will mean for the very skills education is supposed to build. Unless urgent action is taken, RAND warns, schools risk letting AI deepen inequities, anxieties, and mistrust, rather than unlocking its potential to enhance learning and prepare students for the future.