Skip to main content
📰ArticleAI Ethics

When Does AI Assistance Undermine Learning?

AI in Education EditorialUpdated June 2, 20261 min readRead source
When Does AI Assistance Undermine Learning?
🇺🇸US👨‍🎓Students🎯Ethics & Detection🌍Global👩‍🏫Teachers🏛️Administrators+8 more

Using AI as a tutor is like keeping a big jar of cookies in the kitchen cabinet, says Wharton professor Hamsa Bastani . “You tell yourself that you’re just going to eat one, but it’s a slippery slope,” she said. “Self-regulation is hard, even when you know something isn’t good for you.” This insight drives Bastani’s latest research on a critical question in AI-assisted learning: Why do students over-rely on AI help even when they understand it hurts their long-term

Analysis & Perspectives

People Also Ask

How can ChatGPT be used in education?
ChatGPT can serve as an on-demand study partner, essay feedback tool, and concept explainer. Students use it to get plain-language explanations of complex topics, generate practice questions, and receive feedback on drafts. Teachers use ChatGPT to brainstorm lesson ideas, create differentiated materials, and draft parent communications.
Should schools allow students to use ChatGPT?
Leading education researchers and organizations like UNESCO recommend that schools develop clear, transparent policies rather than blanket bans. Allowing ChatGPT for research and brainstorming while requiring original final work teaches students to use AI responsibly — a skill they will need in virtually every career.
Is ChatGPT free for students?
Yes, ChatGPT offers a free tier at chat.openai.com that students can use without a subscription. The free version uses GPT-3.5 and has some usage limits. OpenAI also offers ChatGPT Edu, a discounted plan for academic institutions, with GPT-4o access and data privacy controls required by many schools.
What are the limitations of using ChatGPT for education?
ChatGPT can produce plausible but incorrect facts (hallucinations), lacks real-time information unless connected to browsing, and cannot replace the relational mentorship students need from teachers. It also reflects biases present in its training data, so outputs on sensitive topics require critical evaluation before classroom use.

Related Articles

Making AI chatbots friendly leads to mistakes and support of conspiracy theories
📰 article

Making AI chatbots friendly leads to mistakes and support of conspiracy theories

Skip to main content Skip to navigation Skip to navigation Warm chatbots are 30% less accurate and 40% more likely to support false beliefs, the study found. Photograph: Thai Liang Lim/Getty Images View image in fullscreen Warm chatbots are 30% less accurate and 40% more likely to support false beliefs, the study found. Photograph: Thai Liang Lim/Getty Images Friendly AI chatbots more likely to support conspiracy theories, study finds Chatbots programmed to respond warmly even cast doubts on Apollo moon

Auto-Discovery·Apr 29
AI gives more praise, less criticism to Black students
📰 article

AI gives more praise, less criticism to Black students

Skip to content Credit: Getty Images The Hechinger Report covers one topic: education. Sign up for our newsletters to have stories delivered to your inbox. Consider becoming a member to support our nonprofit journalism. Get important education news and analysis delivered straight to your inbox Choose as many as you want (Required) Weekly Update Elementary to High School Higher Education Proof Points Climate Change Early Childhood Email (Required) Δ As schools introduce artificial intelligence into the classroom, a new analysis

Auto-Discovery·Apr 27
Generative AI Makes Good Research Better, But Demands Human Discipline
📰 article

Generative AI Makes Good Research Better, But Demands Human Discipline

Skip to main content Skip to search YU News Search Offset Search Offset News Channel YU News Featured News Faculty News Schools & Depts Alumni Katz Lieberman Revel Straus Sy Syms Wurzweiler Libraries News Press Releases Publications Contact Alumni Katz Lieberman Revel Straus Sy Syms Wurzweiler Libraries News Generative AI Makes Good Research Better, But Demands Human Discipline Apr 24, 2026 By: daviddefusco Sy Syms Assistant Professor Travis Oh is a co-author of the report: New Tools, New Roles: A

Auto-Discovery·Apr 24