Overview

• A USC-led survey of 1,505 teachers across five countries found about 27% use AI weekly for lesson preparation, though roughly half of those seldom use it during live instruction.
• A HEPI survey reported UK undergraduate use jumped from 66% to 92% in a year, with students turning to AI to explain concepts, summarize articles and suggest research ideas, and 18% inserting AI text into assignments.
• Stanford researchers say overall cheating levels look similar to pre-AI patterns, yet tactics have shifted from copying peers to using AI for tasks such as drafting papers or summarizing books.
• Educators increasingly cite time savings from AI alongside worries about plagiarism and reduced creativity, reflecting a push to set clearer classroom expectations for ethical use.
• Policy responses now emphasize managed use: New York City reversed an early ChatGPT ban, and Los Angeles Unified requires training, parent-student pledges and bars generative tools for students under 13, while teachers pilot in-class writing, device lockdowns, oral checks and portfolio work.