Lawmaker: Education bill’s not an ‘AI ban’ — but it’s not a ‘blank check,’ either

Sen. John Johnson, R-North Ogden, insists his classroom technology bill being considered by the 2026 Utah Legislature is not an AI ban.

But it’s not a “blank check,” either.

Of course, some are alarmed by artificial intelligence in schools. And there appear to be legitimate factors justifying such emotions. Studies warn AI in K-12 classrooms can have a negative effect on children’ s ability to think for themselves — while also affecting a young students’ social or emotional development.

Meanwhile, a 2025 Common Sense Media report, also found that about 1 in 3 teens have used AI companions for social interactions and relationships — including role-playing, romantic interactions, emotional support, friendship or conversation practice.

But when utilized effectively and responsibly, AI has also shown it can be useful in, say, helping students learn to read and write — especially for students learning a second language.

And teachers can also utilize AI to automate some job tasks — allowing them more time to focus on their students.

While presenting Senate Bill 322 that calls for creating an “educational technology regulatory sandbox” in Utah’s public education system, Johnson acknowledged this reality: AI is already in Utah classrooms. Students and teachers are using AI. And vendors are selling AI.

Without structured frameworks, warned Johnson, the state’s public classrooms face several risks.

First, school districts might adopt AI or software tools with no safety vetting — “exposing students to inappropriate content, data harvesting or systems that simulate emotional relationships with minors.”

And those sorts of exposures, he added, may then prompt lawmakers to react “with blunt prohibition,” nixing legitimate innovation that could, say, assist struggling readers or students with disabilities.

And, added Johnson, there’s a third risk: Statewide adoption could happen through administrative mandate rather than legislative authorization — “bypassing the deliberative process this (legislative) chamber exists to provide.”

SB322, said the bill sponsor, is designed to help mitigate such risks by creating Utah’s Education Technology Regulatory Sandbox.

The “sandbox,” he explained, would provide “a structured, evidence-based, evidence-driven framework for evaluating AI in public education before any statewide commitment is made.”

What would Utah’s educational AI sandbox look like?

During presentations to lawmakers, Johnson said SB322 is grounded in balance.

It doesn’t seek to prohibit AI. Instead, he said, it’s designed to discover optimal ways to manage AI.

“Sandbox” is a term often used in the business world to describe a controlled environment where companies can test new technologies or models before implementing them on a larger scale.

The AI education “sandbox” proposed by SB322 would function as a time-limited pilot program that would allow schools to test AI tools under real classroom conditions — while permitting the state to monitor results.

“Participation is entirely voluntary for schools, educators, parents and students,” said Johnson.

Before any AI system can participate in the pilot program, vendors would be required to complete “red teaming” — an adversarial testing practice designed to, say, expose what happens when a child tries to manipulate system confines.

Meanwhile, added Johnson, parental “opt-in/opt-out” authority is assured throughout the program.

“No AI system under this bill can assign final grades, make placement decisions, or override a teacher’s professional judgment without human review and approval,” he said.

“The ‘human in the loop’ requirement is not a suggestion, it is a condition of pilot authorization. Educators must review and approve before any AI output affects a student’s official record.”

The bill also requires legislative involvement in any statewide authorization.

SB322, noted the bill sponsor, extends beyond simply protecting students from “bad technology.”

“It establishes a student ‘due process and dignity framework’,” said Johnson. “AI systems are prohibited from simulating romantic or personal relationships with students”

The proposed AI sandbox program, he added, would also offer a “safe harbor” that provides responsible research and evaluation, while protecting individual privacies.

Meanwhile, students would be entitled to know when they are interacting with AI, and they can request human-educator review of AI-generated decisions before any take final effect.

Additionally, they can never be penalized for declining AI tools.

“This bill does not ask us to trust that AI works,” said Johnson. “It requires proof.”

Before any statewide adoption, independent evaluators would also be required to perform pre- and post-tests on the students to make sure that these systems are useful and actually producing results.

Balancing classroom AI ‘innovation’ and ‘protections’

Johnson said the Beehive State already enjoys a history of leading the nation in balanced technology policies — including the country’s first State Office of Artificial Intelligence Policy.

“We have demonstrated repeatedly that we can write AI policy that respects both the promise of new technology and the constitutional liberties of people it affects,” he said.

When addressing legislative committee concerns about protecting students, Johnson pointed to the proven value of responsible, regulated AI educational tools. “We have to move away from this idea that the sky is falling and everybody’s going to die because some AI bots are going to take over the world.”

AI is obviously here to stay.

“So responsibly,” said Johnson, “we need to put up guardrails for it; but I don’t believe we really want to ban it entirely.”

The Senate approved SB322. It now awaits a final vote from the House.