Idaho approved K-12 AI standards in August 2026. What do they say?

The Idaho State Board of Education approved K-12 AI standards on 12 August 2026. They now go to the Legislature for approval in the next session and would take effect in the 2027-28 school year. State Superintendent Debbie Critchfield was explicit about what they are not: "There's nothing in it that compels anyone to utilize AI, at any level, for any content area."

That makes Idaho the first state with an approved set of K-12 AI standards, and the most detailed statutory picture in the country. It is worth understanding even if you do not live there, because the language is specific enough that other states will borrow it.

Where the standards come from

Senate Bill 1227 took effect on 1 July 2026, adding sections 33-7001 to 33-7008 to Idaho Code. It directs the Department of Education, in consultation with educators, industry partners and technology experts, to develop and recommend:

Generative AI literacy standards for K-12 students, including the knowledge and skills required to understand what generative AI is, how it works, its appropriate and age-appropriate uses, and how to use it ethically, securely, and transparently.

The statute also requires assessment guidelines and a professional development plan, mandates a statewide framework prioritizing "human-centered oversight, transparency, safety, and data security," and states plainly that no generative AI may replace or eliminate a human teacher.

What the standards ask students to do

The drafts that went forward split K-8 and 9-12. The K-8 technology standards carry an AI category alongside computational thinking and digital literacy, with strands on AI systems, AI privacy and safety, and AI literacy. The grade 6-8 statements are the most quotable, and they are notably unsentimental about the technology:

Students should explain how people "design, train, improve, and monitor AI systems" and "why humans remain responsible for important decisions and outcomes." They should analyze how "the variety and quality of training data can impact the functionality and outcomes of an AI system." They should evaluate AI tools "for fairness, accessibility, transparency, explainability, privacy, safety, intellectual property, and accuracy, including how they may help and harm different groups." They should recognize "that AI lacks sentience, emotions, and personal beliefs."

And — the standard most curriculum will struggle with — they should "challenge AI models by purposefully experimenting with complex prompts to discover their limitations, recognize potential errors, and understand when outputs are unreliable."

The 9-12 standards are organized into three domains: AI Fluency, AI Impact, and AI Technical Aptitude, with sub-domains covering appropriate use, college and career readiness, evaluation and critical thinking, ethics and privacy, impact on society, data and analysis, and AI methods.

The definition worth stealing

The draft 9-12 glossary defines the term better than most policy documents manage:

AI Literacy represents technical knowledge, durable skills and future-ready attitudes required to thrive in a world influenced by AI. It enables learners to engage, create, manage, and design AI, while critically evaluating its benefits, risks, and ethical implications.

Note "durable skills." Somebody in Idaho understood that a standard built around a specific tool would be obsolete before it was taught.

One detail nobody should skip: it allows unplugged work

Standard 6-8.AI.1.4 asks students to "develop, train, test, and refine a simple and ethical classification, prediction, or generation model using an age-appropriate tool, simulation, or unplugged process." That last phrase is doing real work. The state's own curriculum people concluded that you can teach the mechanism with paper and a table of examples, which is good news for every school without a device budget and every co-op meeting in a church hall.

What is still uncertain

Legislative approval has not happened. Idaho standards go to the Legislature, and a set of standards can be sent back. The 2027-28 effective date assumes approval in the coming session.

The framework is also, by the superintendent's own description, largely recommendations — teacher training, age-appropriate instruction, and model policy language for districts. Idaho has not told any teacher they must use AI in class, and has been careful to say so.

What to do with it now

If you teach in Idaho: the draft statements are the best available preview of what will be assessed, and the department's AI page is where the final versions will appear.

If you do not: use them as a specification. The Idaho grade 6-8 statements are the clearest public answer to "what should a thirteen-year-old actually be able to do about AI," and any course you are considering can be held against them line by line.

ZEROTH publishes a crosswalk against the Idaho draft codes and the Utah statutory concepts. We describe it as "aligned to" and "prepares for," never as meeting a requirement — ZEROTH is independent curriculum, not accredited, and not a state-approved course. When Idaho's standards clear the Legislature, the crosswalk gets revised against the final text.