Proposed computer-science standards would teach young children that AI is made by people, is not human and can be wrong. Older students would study bias, misinformation, privacy and how reliance on artificial intelligence can affect their own thinking.
By The Redemption Project Newsroom
Education / Technology / Systems Explained
NASHVILLE, Tenn. — Tennessee is considering a new approach to artificial intelligence in public schools that begins with an unusually simple lesson.
AI is made by people.
It is not human.
And it can be wrong.
Those ideas appear in proposed revisions to Tennessee’s K-12 computer-science standards that the State Board of Education will discuss during a workshop Thursday and consider on first reading Friday. If ultimately approved, the revised standards are intended for the 2027-28 school year.
The proposal does not simply teach students how to use artificial intelligence. It repeatedly asks students to evaluate it.
That may be the more important change.
AI skepticism begins in K-2
The proposed kindergarten-through-second-grade standards say students should recognize that AI tools are created by people to help perform tasks and that artificial intelligence “is not human and can sometimes make mistakes.”
That does not mean Tennessee plans to put a chatbot in front of every kindergartner.
The standards say elementary computer science can be taught through stories, play, movement, discussion, manipulatives and activities that do not require computing devices. The state also says the standards are not intended to mandate a particular platform, programming language or technology product.
By grades 3-5, the expectations become more explicit. Students would distinguish human intelligence from artificial intelligence, recognize that AI output should be reviewed by people and learn that the quality of input data affects predictions and responses.
Middle-school students would be expected to evaluate AI output for accuracy, bias, limitations and misinformation and consider effects on privacy, fairness, education and careers.
At the high-school level, the proposed standards go further into reliability, ethics, social effects, environmental impacts and the consequences of relying on AI for learning and productivity.
That progression reflects a notable philosophy.
The state is not treating AI output as authority.
It is treating AI as a tool students should learn to question.
Tennessee’s computer-science program has already grown quickly
The proposal arrives after several years of rapid expansion in computer-science education.
The Department of Education says participation in computer-science courses among students in grades 6-12 increased from 46,679 in 2022 to 158,149 in 2026. It says the number of schools offering those courses increased from 285 to 780 during the same period and that more than 2,000 educators have earned a computer-science endorsement.
Those are state-reported participation figures. The workshop presentation does not specify in the cited slide whether the student totals represent unique students across every measure, so the figures should not be used to calculate a statewide participation rate without additional clarification.
The larger trend is clear.
Computer science has moved from a specialized elective into a much broader part of Tennessee education.
Artificial intelligence is now forcing the state to decide what that instruction should mean.
Standards are easier to write than classroom rules
The proposed standards answer what students should learn.
They do not answer every question about how teachers will teach it.
The Department of Education says future work may include AI policy templates, a statewide AI-literacy framework, crosswalks between AI and other content standards, AI educator credentials and AI standards for teacher-preparation programs. It also says professional learning will be offered to support implementation.
Those plans reveal the implementation gap.
If a middle-school teacher is supposed to teach students how to identify AI bias, what training does that teacher receive first? If a high-school student can use AI as a collaborative tool, which tools may be used and what student information may be entered into them?
If a student uses AI to revise an essay, debug computer code or solve part of an assignment, when does assistance become academic dishonesty?
Those questions are not reasons to avoid AI education.
They are the practical rules schools need before the standards become expectations.
Privacy may be the hardest classroom issue
AI systems can collect much more than the text of a school assignment.
Depending on the product, they can process student prompts, account information, writing samples, uploaded documents and other data.
The proposed standards themselves teach students to think about privacy and security. State and local policy will have to do the same.
Districts need to know whether students may create accounts with commercial AI systems, what vendors may retain, whether parent consent is required, whether student prompts can be used for model improvement and how schools will respond if an AI vendor experiences a breach.
The standards do not prescribe a specific product, which gives local districts flexibility.
It also means different districts could adopt very different rules.
A statewide model policy would reduce the chances that the meaning of student privacy depends on which school system happens to buy the software.
Teacher preparation is the real test
Tennessee’s proposed standards contain a sound instinct: students should learn early that a machine can sound confident and still be wrong.
The harder task is preparing educators to teach that principle consistently.
Elementary teachers were not necessarily trained as computer scientists. Many middle- and high-school teachers entered the profession before generative AI existed in its current form. The technology itself will continue changing faster than a normal standards-revision cycle.
The state’s workshop materials acknowledge that future work must include professional learning, policy guidance and teacher preparation.
That work should accompany the standards rather than trail them.
The State Board has not yet given final approval. Thursday’s workshop precedes first reading Friday, and additional action would follow before the proposed 2027-28 implementation.
The concept, however, is already clear.
Tennessee does not want students merely to know how to ask AI a question.
It wants them to know when not to trust the answer.
I am a retired detective and criminal justice / government educator based in Tennessee. I founded The Redemption Project, as a place to focus on civics, rebuild non-partisan trust, and provide educational and emotional grace while learning about the news. I also have a column in Knox TN Today. My reporting and commentary have also appeared in other outlets including; Governing, The Arizona Capitol Times, South Florida Sun Sentinel, Police1, among other state and regional outlets.
