Issue No. 1 — a periodic brief, not an evergreen page. This field moves fast enough that a brief from six months ago reads as dated; check the verification date before you cite anything below.
Superintendents are, on the whole, more optimistic about AI than teachers — but that optimism is concentrated in back-office use: scheduling, IEP paperwork, differentiated lesson generation, parent communication. On student-facing AI, they're markedly more cautious. Their top fears, roughly in order: student data privacy and compliance exposure under FERPA and COPPA, the risk of being sold free tools now only to be charged once schools depend on them, and a widening equity gap between well-resourced and under-resourced schools. Cheating comes up too, but data governance and cost concerns surface just as often, if not more.
Teachers hold real optimism and real fear simultaneously, often about the same tool. One 2025 survey found 78% want AI to help with grading, lesson differentiation, and daily workload. At the same time, 70–89% worry AI is weakening students' critical thinking and creating dependency on technology for basic tasks. The sharper number: a Gallup/Walton Family Foundation survey of 2,069 teachers (Feb–Mar 2026) found 82% have received no formal guidance on how to use AI, and roughly half of teachers, principals, and district leaders say their school or district has no AI policy at all. The fear is less "AI is bad" and more "I'm being asked to manage this with no training and no rules."
84% of high schoolers report already using generative AI for schoolwork. Most don't consider that cheating: a 2025 survey of 1,000 Americans found about 20% of students describe their own AI use as clearly crossing into cheating, with another 25% placing it in an ethical gray area. The more surprising finding: a separate international survey of 2,373 students found 60% reported real stress related to AI and schoolwork, and a UK-focused breakdown of that research found three in four students worried their own unassisted work would be wrongly flagged as AI-generated. That worry has a documented basis — independent detector testing shows accuracy ranging from the mid-50s to upper-90s percent depending on the tool and text type, and detectors have been shown to misclassify non-native English writers' essays as AI-generated at more than triple the rate of native speakers'.
"Neurodivergent" isn't a fear-and-hopes bucket of its own — autism, ADHD, and dyslexia each carry distinct concerns from parents and distinct hopes from students. See Know Your Student for the breakdown rather than a flattened summary here.
Each brief in this series gets a fresh verification pass before it's marked current — figures aren't left standing just because they were true when first published. If you're citing this brief more than a few months after the date above, check the blog index for a newer issue first.
We're setting up moderated, hold-for-approval commenting — no account required to post, everything reviewed before it appears. Check back soon.