On Thursday, January 15, 2026, a student at Jordan High School in Sandy wrote a paper discussing the implications of artificial intelligence (AI) in education.
Key Points
- The proposed education bill aims to optimise AI usage in classrooms while prioritising student safety.
- Bill sponsor, Senator John Johnson, emphasises that the legislation does not constitute an AI ban.
- The initiative includes a voluntary technology “sandbox” that provides a structured framework for evaluating AI in Utah schools.
Senator John Johnson, a Republican from North Ogden, has clarified that his classroom technology bill under consideration by the 2026 Utah Legislature does not intend to implement a ban on AI. However, he also asserts it is not a “blank cheque” for unrestricted AI use.
Concerns regarding the presence of AI in educational settings are valid, as studies indicate that AI could negatively impact students’ critical thinking skills and social-emotional development. A report from Common Sense Media in 2025 highlighted that approximately one in three teenagers have engaged with AI companions for social interactions, ranging from role-playing to emotional support.
Despite these concerns, when employed judiciously, AI has demonstrated potential benefits, such as aiding students in learning to read and write, particularly for those acquiring a second language. Additionally, teachers can leverage AI to automate administrative tasks, thereby dedicating more time to student engagement.
Senate Bill 322 proposes the establishment of an “educational technology regulatory sandbox” within Utah’s public education system. Senator Johnson acknowledged that AI is already a reality in classrooms, with both students and educators utilising various AI technologies.
Without a structured framework, Johnson warns that public classrooms may face significant risks, including the adoption of unvetted AI tools that could expose students to inappropriate content or data exploitation. Such scenarios could lead lawmakers to respond with severe restrictions, potentially stifling beneficial innovations that could assist struggling students.
Moreover, Johnson expressed concern that the adoption of AI tools might occur through administrative measures rather than through the legislative process, undermining the thorough deliberation that such decisions warrant.
SB322 aims to mitigate these risks by creating a regulatory sandbox for educational technology. Johnson stated that this sandbox would serve as a structured, evidence-based framework for evaluating AI’s role in education prior to any broader implementation.
What will Utah’s educational AI sandbox entail? During discussions with lawmakers, Johnson emphasised that SB322 seeks balance rather than prohibition. The sandbox, a concept commonly used in the business sector, will allow educational institutions to trial AI tools in real classroom environments while enabling the state to monitor the outcomes.
Participation in this pilot programme will be entirely voluntary for schools, educators, parents, and students. Prior to engaging with any AI system, vendors must undergo “red teaming” — a testing practice aimed at identifying potential vulnerabilities.
Parental consent will be a cornerstone of the programme, ensuring that no AI system can make final decisions regarding grades or student placements without human oversight. Johnson reinforced that the “human in the loop” principle is a mandatory condition for the pilot’s approval.
Furthermore, the bill mandates legislative involvement in any future statewide authorisation of AI tools. Johnson noted that SB322 is not solely about shielding students from harmful technologies; it also establishes a framework for student due process and dignity, explicitly prohibiting AI systems from simulating personal relationships with students.
The proposed AI sandbox would facilitate responsible research and evaluation while safeguarding individual privacy. Students will have the right to know when they are interacting with AI and can request human review of AI-generated decisions.
Importantly, students will not face penalties for opting out of AI tools. Johnson stated, “This bill does not ask us to trust that AI works; it requires proof.” Independent evaluators will conduct pre- and post-tests to ascertain the efficacy of these systems before any statewide implementation.
In addressing concerns about student safety, Johnson highlighted Utah’s history of leading in balanced technology policies, including the establishment of the first State Office of Artificial Intelligence Policy in the nation. “We have consistently demonstrated our ability to formulate AI policies that honour both the potential of new technologies and the constitutional rights of individuals,” he remarked.
As AI becomes an integral part of education, it is imperative to navigate its integration with caution and foresight.