The US Department of Education published its final supplemental priority on “Advancing Artificial Intelligence in Education,” establishing a framework that will shape how federal discretionary grant dollars flow toward AI-related projects across K-12 and higher education. For education leaders, this is the clearest signal of where the McMahon Department wants to steer innovation funding—and what it deliberately chose not to do. 

First, some important context: This is not a regulation, a mandate, or a new program. A supplemental priority is a policy tool the Secretary can attach to existing discretionary grant competitions as an absolute, competitive preference, or invitational priority. This priority joins a growing group of supplemental priorities the Department has rolled out since September 2025, covering evidence-based literacy, educational choice, career pathways, and other administration themes.

Over 300 Comments And What They Reveal

The proposed priority drew more than 300 public comments, reflecting a genuine tension in the field. Supporters praised the Department for recognizing AI’s workforce implications and its potential to personalize learning. Opponents raised concerns about screen time, cognitive development, student privacy, and the absence of rigorous evidence on AI’s impact in classrooms. The Department acknowledged these concerns but largely held its ground, arguing that the United States must prepare students to compete in an AI-driven economy and that many decisions about safety are best made at the state and local level. 

Commenters explicitly asked for vendor transparency on AI model training, parental consent mandates, and cybersecurity standards. The Department declined all three, arguing these questions are “optimally decided at the state and local level.” 

K-12: Literacy, Professional Development, and Integration

The K-12 provisions span AI literacy instruction (including detecting AI-generated misinformation), expanded computer science and AI course offerings, educator professional development, dual-enrollment pathways, and career-relevant certifications. On the use side, the priority encourages AI for personalized and differentiated instruction, high-impact tutoring, and college and career navigation, along with explicit emphasis on supporting students below grade level, those needing remediation, and students with disabilities. A new provision on universal design for learning signals that accessibility should be embedded in AI projects from the start.

Two things stand out. First, the priority references “high-quality instructional resources” but makes no mention of high-quality instructional materials (HQIM), which is a surprising omission given that the Department’s Meaningful Learning Opportunities priority, finalized just two months earlier, treats HQIM as a central framework.

Second, the Department strengthened language around age-appropriate use and added a provision on training educators in age-appropriate AI methodologies, but declined to define what “age-appropriate” actually means, deferring to states, districts, and families. That’s flexibility for educators and a fragmented landscape for the frontier labs trying to build for K-12.

Higher Education: Preservice Training and Curriculum

The higher education provisions are narrower but important. The priority supports embedding AI and computer science into education curricula and into preservice and in-service teacher preparation programs. 

Definitions Worth Watching

The priority defines three terms. AI adopts the statutory definition of a machine-based system that makes predictions, recommendations, or decisions. Computer science spans hardware and software design, computational thinking, coding, analytics, machine learning, and AI. 

AI literacy is the “technical knowledge, durable skills, civic awareness and future-ready attitudes”—ethical reasoning, critical social inquiry, interdisciplinary problem-solving, creativity—needed to engage with, create, manage, and design AI while evaluating its benefits and risks. That’s a challenge in itself. Phrases like “durable skills,” “future-ready attitudes,” and “civic awareness” lack any settled meaning or consensus.

The Bottom Line

To the Department’s credit, the priority is thoughtful. It frames AI as a means rather than an end, centers teaching and learning over technology for its own sake, and adds meaningful language on universal design and age-appropriate use. 

But sidestepping parental consent may be a decision the Department comes to regret. The backlash against AI and screens in schools is building, not receding. Twelve state bills have been introduced in 2026 alone addressing concerns about replacing teacher-led instruction with digital learning tools. And just this week, New York City withdrew plans to open Next Generation Technology High School, the city’s first AI-focused public high school, following sustained opposition from parents. 

The other challenge is that the document establishes policies for an earlier era of ChatGPT-4 chatbots and classroom tools, not for the agentic systems that frontier labs are introducing and that are likely to enter the education system later this year. Agentic AI doesn’t just answer questions but acts autonomously, makes decisions, and works for hours to accomplish a wide range of tasks. The governance questions those systems raise about autonomy, accountability, consent, privacy, liability, and the appropriate scope of AI decision-making.

The Department has outlined thoughtful priorities. But the technology is evolving faster than any priority document can keep up with, and school leaders are already wrestling with it in real time.