AI cheating is endemic in high schools and colleges. Stories proliferate showing just how many students unashamedly use AI programs like ChatGPT and Gemini to do their academic work for them, whether that's completing math homework or writing entire essays. But why is this happening? Here's one underrated factor: many students may be turning to AI because they can't handle the academic rigors of college, and are turning to large language models like ChatGPT to hide their deficiencies. That, mixed with old-fashioned laziness, could be driving the proliferation of AI cheating. For both situations, the solution may be less college education, not more.
What does laziness mean? Well, plenty of students who are clearly capable of handling college-level work still use ChatGPT and other AI models to do their school work. In one recent article, Hua Hsu, a staff writer for The New Yorker, interviewed several NYU students who openly admitted they use AI to cheat on assignments.
"We had to read Robert Wedderburn for a class," one student told Hsu, (Wedderburn was a 19th-century Jamaican abolitionist). "But, obviously, I wasn't tryin' to read that." Instead, the student told Claude, a large language model developed by Anthropic, to turn the reading into "concise bullet points."
