85% of Teachers Use AI. Only 18% Got Any Training on How to Use It.

By Hector Herrera | November 13, 2026 | Education

Generative AI is already embedded in most US classrooms — but the people running those classrooms were largely left to figure it out alone. Google's 2026 AI in Education report found that 85% of educators regularly use generative AI tools, while only 18% of US teachers received formal training or guidance on responsible use. That 67-point gap is not a technology adoption problem. It's a professional development failure at institutional scale.

The practical result: teachers are using AI every day — to generate lesson plans, write assessments, differentiate instruction, and manage administrative load — without shared standards, ethical frameworks, or consistent guidance on where the tools help and where they mislead. That's a risk for students, and it's a liability for school districts.

The Numbers in Context

Google surveyed educators across the US for this report, capturing adoption patterns and training experiences across K-12 and higher education settings.

Key findings:

• 85% of educators regularly use generative AI tools in their work
• Only 18% received formal training or institutional guidance on responsible AI use
• The majority of AI-using teachers are self-taught — piecing together knowledge from YouTube, social media, and peer conversation
• The gap is widest at the K-12 level, where district-level AI policy is uneven and professional development budgets are constrained

For context: when internet access rolled into schools in the late 1990s, it took more than a decade to develop e-safety curricula, digital literacy frameworks, and acceptable use policies. AI is moving faster than that — and the evidence suggests institutions are further behind.

Why the Gap Is Dangerous

Untrained AI use in classrooms creates risks that fall into three categories:

Accuracy and misinformation. Generative AI tools hallucinate — they produce confident-sounding text that is factually wrong. Teachers without training on this failure mode may not recognize or flag AI-generated errors before they reach students. A teacher using AI to generate a history timeline or a science explainer who hasn't learned to verify outputs is a potential misinformation vector inside a trusted institution.

Equity and access. AI tool access is not evenly distributed. Schools with better-resourced technology programs, more tech-confident staff, and institutional subscriptions to premium AI tools are pulling ahead of under-resourced schools where teachers are using free-tier tools — or no tools at all. Without coordinated professional development, the AI fluency gap between high- and low-resource schools will widen.

Academic integrity. Teachers who don't understand what AI can do also can't effectively set policies about what students are and aren't permitted to do. A teacher who doesn't understand AI-generated text can't detect it, evaluate it, or teach students how to use it responsibly. The result is policy vacuum — schools with no coherent position on student AI use because the adults don't yet have one on their own.

What Responsible AI Training Actually Requires

Formal AI guidance for educators isn't just a few hours of professional development. The frameworks that have shown results — including those developed by ISTE and piloted in forward-looking districts — involve:

• Tool literacy: how generative AI works, what hallucination is, and how to verify AI outputs
• Pedagogical frameworks: when AI use supports learning versus shortcuts it
• Student-facing policy: how to set clear, enforceable expectations for student AI use
• Ethical dimensions: bias in AI systems, data privacy, and intellectual property

None of this is complicated, but all of it requires deliberate instruction — which 82% of US teachers haven't received.

The Institutional Response Is Lagging

The Google report lands as school districts are still sorting out basic policy questions. Some districts have banned AI tools outright. Others have no policy at all. A smaller number — mostly in well-resourced suburban districts or forward-thinking urban systems — have developed thoughtful AI use frameworks with accompanying professional development.

The absence of federal guidance compounds the problem. The Department of Education has published resources on AI in education, but there is no binding national framework requiring schools to train teachers on AI tools before deploying them — or even disclosing which tools are in active use.

State-level action is uneven. A handful of states, including California and New York, have begun developing AI literacy standards, but none have made teacher AI training mandatory. Without a mandate, professional development competes for the same constrained budgets as everything else schools are trying to do.

What to Watch

The Google report will add pressure on states and districts to develop formal AI guidance requirements — particularly as the 2026-2027 school year deepens AI integration. Watch for whether the Department of Education moves to formalize AI training standards, and whether any state legislature introduces mandatory professional development requirements for teacher AI use. The parallel story is vendor-side: Google, Microsoft, and other platform providers have an incentive to fill the training gap their own tools are creating.