Education is a sensitive AI deployment context — student safety, academic integrity, and equity considerations create constraints other industries do not face. But the administrative load on teachers is also brutal. Claude can help teachers; it should not interact with students in most contexts. Here is the framework.
1. Lesson planning. Curriculum aligned to standards, differentiated for skill levels.
2. Assessment design. Quizzes, rubrics, project assignments.
3. Parent communications. Conference prep notes, behavior reports, progress updates.
4. Grading support. Rubric application, feedback drafting on student work.
5. IEP and 504 documentation. Drafts of accommodations and documentation.
Direct student tutoring without teacher oversight. Tutoring AI products exist; deployment requires careful framework.
Counseling or wellness conversations. Human counselors only.
Grading or evaluation of student work without teacher verification. Bias risk is real.
Direct messaging students. Authentication and safety requirements exceed what general AI provides.
FERPA compliance. Student educational records cannot enter AI tools without proper safeguards.
COPPA for K-12. Special restrictions for under-13 students.
District AI policy. Most districts now require written AI usage policies for teachers.