Screen time and artificial intelligence (AI) are the hottest topics of conversation in education policymaking circles around the country. It’s the latest evolution of the ed sector’s “techlash” that grew from concerns about the impact of social media on kids and from states enacting policies to limit or ban student cellphone use in K-12 schools. Nearly 90% of tweens and teens across the country are interacting with AI, while 80% of K-12 educators say they use AI-powered tools. So, it’s no surprise that this rapid shift is contributing to concerns about the use of technology in schools. Thoughtful guardrails are appropriate, but blunt policy responses will overcorrect and leave teachers wrestling with the practical realities of AI implementation.

The most visible policy responses have been bans and limits, although their design varies. Screen-time policies generally impose the strictest restrictions on younger students and allow more use as students get older, often with exceptions for assistive technology (e.g., screen readers for students with visual impairments). AI policies are less consistent: New York City prohibits student-facing AI through eighth grade while permitting limited, supervised use in high school; Los Angeles Unified School District’s AI moratorium applies across grade levels but only to district-managed devices and accounts. Both districts’ policies avoid addressing an arguably larger challenge: student screen time and AI use outside of school.

But what kind of action is helpful for students and educators? 

At a surface level, screen time and AI restriction policies appear straightforward. But dig a little deeper and they often obscure nuances that matter for teaching and learning. If schools remove digital tools, what ensures their replacements will offer something better? If use is limited by minutes, who’s responsible for tracking and enforcing those limits? And do policies distinguish between thoughtfully developed ed tech tools and those that erode students’ productive struggle? Policies that don’t address those questions obscure thoughtful solutions and leave teachers and students to deal with the consequences.

This creates a difficult situation for school leaders. Do they adopt broad restrictions on screen time and AI, knowing that some of the messy details of implementation will fall on the shoulders of teachers and students? The bad news: there’s no playbook with the “right” answer for every school navigating the techlash. The good news: there are a few questions K-12 leaders can ask as their communities recalibrate how technology is deployed in classrooms:

• What’s the current state of students’ screen time and AI use? Before heading down the path of adopting restrictions on screen time or AI, leaders must first understand how much time students spend interacting with digital tools. Understanding the contours (and limits) of current data can help clarify the scope of what policies are needed. It might make sense to start with getting better data!

• What are leaders trying to protect or improve? Assuming there’s a gap between the current state and what a school community desires for screen time and AI use, how big is the delta? How do those expectations differ by grade level? Are some uses of screen time and AI preferred to others, and if so, how will leaders differentiate among them? If screen time limits are adopted, will some subjects (such as math and reading) get priority, or should all content areas have a proportional “claim” to digital/AI tooling? Not every one of those details will need to be codified in policy, but they should be part of leaders’ conversations before sweeping policies are enacted.

• What will replace restricted tools? Removal of technology is a much easier matter than figuring out what takes its place. What supports will drive high-quality instruction that is less tech-focused? If restrictions on screen time and student-facing AI don’t come from a coherent effort to support consistent, quality instruction, what’s to stop screen time from being replaced with pencil-and-paper work on AI-generated worksheets?

Responses to legitimate concerns about the use of technology in schools need to offer more than easily quantifiable restrictions on students and teachers. A recalibration of screen time and student AI use should be part of a broader strategy to support higher-quality instruction and better learning outcomes. Otherwise, we may end up with a suite of policies that are better at producing headlines than supporting stronger classrooms.

Quick Hits

• All the small (implementation) things: Recent analyses from the Center on Reinventing Public Education and Chalkbeat, along with Bellwether’s updated Learning Systems series, find that while states are providing guidance on AI, districts and teachers need more support to help students navigate using the tools.* Let’s hope that addressing educators’ need for greater support on AI isn’t lost in the conversation about tech restrictions in schools. 

• Gemini’s out of the bottle: Google recently added its Gemini chatbot to Google Classroom for K-12 students. The ability to enable Gemini access for all grades K-12 is distinct from the approach of other large language model providers, which either offer no access for younger students (Microsoft Copilot) or focus enterprise accounts on teachers instead of students (OpenAI and Anthropic). Common Sense Media notes that regardless of school’s decision to allow or restrict Gemini in Google Classroom, it’s not possible for school administrators or parents to disable Gemini features in a basic Google search. 

• Time for AI policy to meet the parents: AI is already shaping how students learn, socialize, and seek support, but families are often only brought into schoolwide AI decisions after a tool has been adopted or a policy has been written. A new report from my Bellwether colleagues finds that in national surveys, 57% of parents have not been asked for input on AI use in schools, and 47% have never received information about their district’s AI policy. Parent engagement should be part of AI governance, not an afterthought — this report offers seven early principles for how districts and advocates can make that happen.

Odds & Ends