The Stanford HAI 2026 AI Index reports that just 6% of US teachers describe their school’s AI policies as clear and comprehensive. The same report finds that only half of middle and high schools have any AI policy in place at all, and of those that do, only 28% permit AI use in some circumstances. Although 30 US states had issued AI guidance for schools as of January 2026, the guidance is largely non-binding and implementation is devolved to local education agencies, producing wide variation in quality and coverage. The data underscores the gap between high-profile state AI education bills and teachers’ day-to-day experience of policy ambiguity.
Who it affects: US public K-12 teachers; school administrators; state education agencies; parents and students navigating inconsistent classroom AI rules.
What is notably missing: The 2026 AI Index describes variance but does not publish a standardised rubric states could adopt to harden guidance into binding policy. No federal framework mandates that states move from guidance to enforceable standards. The survey does not break out whether teachers in states with the longest-standing guidance (e.g. Ohio, California) perceive greater clarity than those in later-moving states, which would be useful for benchmarking.