![\"Stanford's](\"https:\/\/www.europesays.com\/ai\/wp-content\/uploads\/2026\/04\/1777150271_799_0x0.jpg\")
<\/p>\n<\/p>\n
Stanford’s Institute for Human-Centered AI released its 2026 AI Index Report <\/p>\n
getty<\/p>\n
When Stanford’s Institute for Human-Centered AI released its 2026 AI Index Report<\/a> on April 13, the headlines mostly went to the model benchmark battles, the $581.7 billion in global corporate AI investment, and the data center power numbers. Buried in chapter seven, however, is the finding that should be on the desk of every school leader.<\/p>\n Four in five U.S. high school and college students now use AI for schoolwork. Data is similar in the UK. Only 6% of teachers say their school’s AI policies are clear.<\/p>\n That is not a guidance gap. That is an institutional readiness failure and it is happening right now, while the systems built to support teachers continue to lag behind the technology students are already using every day.<\/p>\n A few weeks ago I working with a school on the upper east side of Manhattan and their survey of their middle and hig school students found that 100% were using AI. <\/p>\n The Numbers That Should Reframe Our Efforts<\/p>\n The Stanford report, produced in collaboration with the Kapor Foundation, the Computer Science Teachers Association, and the Expanding Computing Education Pathways Alliance, is now in its ninth edition and remains one of the few comprehensive AI datasets not produced by an organization with a financial stake in the outcome.<\/p>\n Its findings on K-12 are stark. Estimates of student AI use for schoolwork range from 50% to 84%, drawing on data from the College Board, RAND, and the Center for Democracy and Technology<\/a>. High school students most often use generative AI for conducting research, editing essays, and brainstorming – three activities that sit at the heart of how schools have traditionally measured learning.<\/p>\n Yet only about half of middle and high schools have any AI policy at all. Of those that do, just 28% permit AI use in some circumstances, while 22% prohibit it. Forty-seven percent of students said they have wanted to use AI for schoolwork but were unsure whether it was allowed.<\/p>\n A separate, independent survey from RAND released in September 2025<\/a> reached the same destination by a different route. Researchers found that 54% of students and 53% of English language arts, math, and science teachers were using AI for school-related work by spring 2025. That was a jump of more than 15 percentage points in roughly a year. Yet over 80% of students reported that their teachers had not explicitly taught them how to use AI for schoolwork. Only 34% of teachers said their school had policies on AI use related to academic integrity.<\/p>\n When two of the most rigorous independent studies in the field arrive at the same finding, school leaders no longer have the option of waiting for clarity to emerge.<\/p>\n What Most Coverage Got Wrong<\/p>\n Most reporting on the AI Index has framed this as a story about students racing ahead while schools fall behind. That framing is incomplete, and it lets the wrong people off the hook.<\/p>\n The deeper finding in chapter seven is that the responsibility for AI policy is being pushed almost entirely down to local education agencies. As of January 2026, in the US, 30 states have issued some form of AI guidance, but according to research from the Expanding Computing Education Pathways Alliance cited in the report, state-level guidance is largely nonbinding and decentralised. Most states are relying on existing federal frameworks like the Children’s Online Privacy Protection Act and FERPA rather than issuing AI-specific mandates. Only five states have allocated specific professional development funding for AI education.<\/p>\n The April 2025 federal Executive Order 14277, “Advancing Artificial Intelligence Education for American Youth,”<\/a> created a White House AI Education Task Force and a Presidential AI Challenge. But the order’s actual implementation depends on discretionary grants, public-private partnerships, and existing teacher preparation pathways. As the Stanford report notes, AP Computer Science, which is one of the most common pathways to advanced computing in U.S. high schools, does not include AI-specific content.<\/p>\n What that means in practice is a teachers in very different contexts are being asked to figure this out on roughly the same timeline, with vastly different resources. The RAND data shows the consequences. Teachers in higher-poverty schools are less likely to use AI, and principals in the highest-poverty schools are less likely to receive AI guidance. This is a divide that, as RAND researchers warned in coverage of their findings<\/a>, risks turning AI rollout into another driver of educational inequality.<\/p>\n Three Things Educators Can Do Now<\/p>\n Sitting with the data does not change classrooms. Acting on it does. Three concrete moves are within reach of most school leaders before the summer break.<\/p>\n 1. Write a one-page AI policy that students can actually understand. <\/p>\n The Stanford data shows the issue is not just the absence of policy but the absence of clarity. Only 36% of students described their school’s policies as extremely clear. A workable policy answers four questions in plain language: when students may use AI, when they may not, how they should disclose AI use when permitted, and what happens if they get it wrong. Schools that have moved fastest on this have one thing in common, according to the report: they treat AI as something to be taught about, not just something to be governed.<\/p>\n 2. Train teachers on prompting before training them on tools. <\/p>\n RAND’s 2025 panel data found that the fastest-growing AI literacy skill among American workers, measured by LinkedIn profile data, is AI prompting. Yet most school AI training continues to focus on specific products, this might be a particular chatbot, a particular grading tool, a particular lesson planner. Tools change every quarter. Prompting is the underlying competence. Schools that invest in the durable skill rather than the latest product will see the return on professional development hold up across budget cycles.<\/p>\n 3. Replace the cheating conversation with a thinking conversation. <\/p>\n