High School Teacher

AI already does a real share of a teacher's behind-the-scenes work. In a Gallup and Walton Family Foundation survey of 2,232 US public school teachers (spring 2025), six in 10 had used AI for their work that school year, most often to prepare lessons, make worksheets and adapt materials for different students. The 32% who used it weekly estimated it saved them 5.9 hours a week, about six weeks over a school year. AI is also starting to teach. In a Harvard experiment published in Scientific Reports (June 2025), students in an intro physics class learned more from a carefully designed AI tutor at home than from an active-learning class, by 0.73 to 1.3 standard deviations. Those were college students and a single pair of lessons, and the authors warn against AI "entirely supplant[ing]" classroom teaching, but the result shows AI can deliver instruction, not just paperwork.

The job market gives AI's time savings nowhere to go. The Bureau of Labor Statistics projects "little or no change" in high school teacher jobs from 2025 to 2035 (about 2,500 fewer on a base of 1,087,500) because the school-age population and enrollment are expected to shrink and hiring depends on state and local budgets. There are still about 63,500 openings a year, mostly to replace teachers who leave or retire. Our rubric says growing demand can offset automation, and there is no growth here to do that. Our read is that the time AI frees up is more likely to show up over the years as bigger classes or fewer new hires than as more teachers. No source we found shows a school district cutting teachers because of AI yet.

It stays at 3 because running a room of teenagers (keeping order, building relationships, spotting a kid in trouble) needs a person, and every state requires public school teachers to hold a license. What would change the score: districts citing AI to raise class sizes or cut positions would push this to 4. Several budget cycles of heavy AI use with no effect on staffing would support moving it to 2.