With the new school year started, one of the growing challenges facing students, parents, and teachers is how to address the classroom uses of AI, which has become a double-edged sword in the education process.
Artificial intelligence, or AI, is technology that can perform tasks that have in the past required human intelligence. NASA says that AI computer systems “perform complex tasks that are normally done by human reasoning, decision making, creating, etc.”
These systems can operate “without significant human oversight” and “can learn from experience and improve performance when exposed to data sets.”
AI tools in the education sector have taken several forms.
For example, there are chatbots like OpenAI’s ChatGPT that respond to user questions by creating summaries, explanations, written materials, and other products.
In addition, there are AI tutoring systems that offer students step-by-step instruction, practice problems, and assessments of their learning progress.
There are also AI writing assistants such as Grammarly that act like editors for users’ writing, word selection, topic selection, and outline organization.
For teachers, there are AI teaching assistants that help plan lessons and grade assignments.
It is estimated that about a quarter of students regularly use AI tools, although many educators believe the real proportion is much higher.
There are certainly pluses to these AI tools.
The Summit Charter Learning network of charter schools says that “students can benefit from asking AI to clarify information or explain concepts in a new way.” And AI can be “a supplemental tool for research.”
Jonah Davids and Leif Rasmussen, co-founders of the research group DeepAudit, have written that AI will “provide personalized instruction at scale to make homeschooling, micro-schooling, and other alternatives to public schooling more feasible than ever before.”
Yet, for all this potential, there are grave downsides to the AI revolution in education, with the biggest problem being cheating.
Jeremy Adams, a teacher of the year in California and book author, recently wrote, “This year we’re grappling with a Herculean task: how to counter the reality that our students, armed with artificial intelligence technologies that were the stuff of sci-fi just a few years ago, can cheat on virtually any assignment we give them.”
From math to writing, he warned, “You name it, they can use AI to cheat on it.” And the price of over reliance on AI is high.
An MIT study of the impact of AI usage on young people who were asked to write SAT essays found that AI users performed worse than non-AI users “at all levels: neural, linguistic, scoring.” Further, with each essay the subjects were asked to write, the AI users got lazier, eventually just copying and pasting.
If students use AI to write their essays and reports, they will fail to learn more than just how to write.
Meredith Coffey, a senior policy associate at the Fordham Institute and a former high school English teacher, has noted, “learning to write helps students learn to read” and helps them understand “sentence and text structure.” It’s important “that we don’t surrender writing instruction to our AI overlords.”
For teachers using AI to design lesson plans, Heather Peske, president of the National Council on Teacher Quality, warns, “Though well-intentioned, there is a risk the AI will design lessons contrary to research on best instructional methods.” The risk is especially high in reading instruction where, despite the overwhelming evidence backing phonics-based methods, “AI may have trained on outdated and discredited reading approaches.”
Finally, there is AI’s political bias. A study by the United Kingdom-based Centre for Policy Studies concluded that leading AI systems “tend to produce content that, on average, manifests left-of-center policy preferences — in some cases markedly so.”
Despite these red flags, the AI genie cannot be stuffed back into its bottle. Thus, the public must demand transparency in AI usage in schools, and schools should consider pedagogies such as the Socratic method where teachers engage in question-and-answer dialogue with students in class to ensure that students have a deep understanding of subject matter.
Fake short-term achievement based on AI will lead to real long-term losses for students and our nation.
Lance Izumi is senior director of the Center for Education at the Pacific Research Institute. He is the author of the PRI book “The Great Classroom Collapse: Teachers, Students, and Parents Expose the Collapse of Learning in America’s Schools.”



