AI for planning, not classroom pacing
Commentary and guidance suggest using AI tools for teacher efficiency—drafting exemplars, parent notes, organizing materials and generating extension questions—while keeping live instructional regulation and moment‑to‑moment judgment teacher‑led. That distinction was drawn in reactions to federal AI grant priorities and a Fast Company critique of narrow AI personalization in schools. (edsource.org, fastcompany.com)
The new line in school artificial intelligence policy is simple: use it to help teachers prepare, not to let software run the room. (edsource.org) The U.S. Department of Education said this week it will give a new competitive preference to grant applicants that use artificial intelligence to improve teaching and learning, according to an April 15 EdSource report citing K-12 Dive. A July 22, 2025 Dear Colleague letter had already told grantees they could use federal funds for artificial intelligence if the use met legal and program rules. (edsource.org, ed.gov) That federal guidance did not frame artificial intelligence as a teacher substitute. The Education Department’s 2025 letter said schools could use it to “support educators” while not replacing their role, and its grants page separately says artificial intelligence can be used for tasks such as transcription and meeting notes. (ed.gov, ed.gov) A Fast Company essay published April 15 pushed that distinction further. It argued that the dominant pitch for school artificial intelligence has been “personalized learning” through tutors and platforms, but said the better use is “personalized teaching” that helps educators plan lessons, draft examples, and respond to students faster. (fastcompany.com) The argument lands after schools spent years testing screen-heavy models and then watching students struggle with isolation and uneven engagement during and after the coronavirus pandemic. The Fast Company piece says systems that put students alone with software reduce teachers to monitors, while teacher-led classrooms still depend on live judgment, pacing, and relationships. (fastcompany.com) In practice, that means using artificial intelligence for back-office classroom work: drafting parent messages, organizing materials, generating extension questions, and producing model responses a teacher can edit. It does not mean letting a model decide, minute by minute, when a class should speed up, slow down, or change direction. (fastcompany.com, ed.gov) California’s own guidance has taken a broader but related approach. EdSource reported that the California Department of Education updated its guidance in January 2026 to encourage schools to integrate artificial intelligence into teaching, assessment, and educator professional development. (edsource.org) The same EdSource report said the new federal priority does not include a plan to evaluate student data privacy or equitable access in school artificial intelligence use. Keith Krueger, chief executive of the Consortium for School Networking, told K-12 Dive the department should consider a separate funding stream so artificial intelligence does not crowd out other programs. (edsource.org) The money is already moving on the higher education side. In January 2026, the department announced $169 million in Fund for the Improvement of Postsecondary Education awards, including grants tied to the responsible use of artificial intelligence to enhance teaching and learning. (ed.gov) So the fight is no longer over whether schools will use artificial intelligence at all. It is over whether the tool writes the worksheet before class starts, or starts calling the plays after the bell rings. (fastcompany.com, ed.gov)
