Artificial Intelligence (AI) is here to stay, and schools will have to find a strategy to make the most of it as a tool, but also ensure students can think on their own, rather than depending on a machine to think for them. Moreover, given the importance of AI in the technological and scientific sectors in particular, the United States must improve its STEM education. How much do you know about AI and education? Play “Two Truths and a Lie” to find out!
A. AI detectors are incredibly accurate and an excellent tool for teachers to use to detect cheating.
B. Student mental health declined and test scores dropped starting in 2012 due to screens.
C. The United States has fallen behind in mathematics and literacy.
A. LIE! AI detectors often have false positives, so teachers can’t rely on them as consistent, reliable indicators of whether or not a student has cheated. Any assignment in which a computer is used, or could have been used—that is to say, essentially any take-home assignment—is necessarily suspect in the age of AI.
B. TRUTH! American schools must return to technology-free classrooms, except in subjects and cases where technology is explicitly required. There are good reasons to do this even beyond concerns around academic integrity. As social psychologist and researcher Jonathan Haidt has argued, even in the absence of AI, children’s mental health has declined over the past decade or so because of ubiquitous smartphone and technology use, especially in schools. “We were promised [education technology, or EdTech] would revolutionize teaching. It was going to engage students, bring scores up,” he said in a recent interview. “And what happened? That’s exactly when scores started going down. … People saw a big drop and said it was COVID. And yes, there was a drop with COVID—but the decline actually started in 2012. So if kids knew so much less after COVID, it’s not because of the virus. It’s because all education was on screens … the distraction effects are enormous. They swamp any possible benefit [of screens]. That’s the conclusion.”
C. TRUTH! Per the National Assessment of Educational Progress (NAEP) results released by the Nation’s Report Card in 2025, 61% of fourth-grade students and 72% of eighth-grade students are not proficient in math. And these students are not magically growing out of their mathematical illiteracy, either: A recent faculty report at the University of California, San Diego—one of the nation’s top public universities—showed that the share of its first-year students who are not able to meet middle school math standards has grown thirty-fold in the past five years. A full one-eighth of UCSD freshmen are unable to do basic math, many not even meeting elementary school standards.
Bottom Line:
As education expert Robert Pondiscio writes for the American Enterprise Institute, being careful about AI in education is “not an argument for technophobia. It’s an argument for intellectual vigilance.” In the age of AI, the worst thing we could do is allow students to let machines do their thinking for them. That will make us, as a nation, dependent on machines rather than making us competent enough to use machines to the human and national interest.
To learn more, read the Policy Focus: Artificial Intelligence and Education.