Editor’s note: This essay is an entry in Fordham’s 2026 Wonkathon, which asked contributors to answer this question: “What rules should govern the uses of digital technology in schools—and who should set them?” Learn more.
Whether or not anyone wrote it down, every school makes technology policy. Where states and districts haven’t set clear rules, teachers and principals make those decisions themselves, one app at a time, while fielding parent pushback, tight budgets, nonstop vendor marketing, and pressure to prepare students for an uncertain future. Each of those pressures is real, and none of them should outrank student learning.
Student learning is the right guide because the same tool can strengthen or weaken instruction. Teachers trained to use a tool to scaffold text-dependent questions can support students in reading grade-level text. Teachers who don’t receive that training may use the same tool to rewrite the text at a far lower level, stripping out the new vocabulary and knowledge students need to become readers.
We propose that leaders at every level judge each tool by one question: Does it improve the time students spend learning rigorous, rich content with strong teachers? A tool belongs in the classroom only when the answer is yes. To aid in that determination, we propose these four rules and assign each to the system level best positioned to make and enforce it. This approach is strict by design and keeps many tools, including promising new ones, out of regular classroom use until they show meaningful results in a pilot. States should fund rigorous pilots so the field learns which of these tools improve student learning.
1. States should define technology categories by instructional purpose.
Policies that lump cellphones, social media, general-purpose chatbots, and instructional software together as “screens” erase the distinctions leaders need to make decisions grounded in instruction. Each category serves a different purpose, carries different risks, and offers different evidence about how it adds to or detracts from student learning and experience.
States do not have to start from scratch. The FullScale and SETDA leadership brief series already separates educational, instructional, operational, and consumer technology. When the state sets these definitions, a tool approved for instruction in one district does not count as restricted screen time in the next.
Within instructional technology, rules should go one level further and separate:
- Teacher-facing from student-facing tools;
- Proven tools, meaning those with published, independent research meeting ESSA evidence Tiers 1–3 and accuracy data for the students who use them, from unproven ones; and
- Tools that support teacher-student interaction from tools that stand in for it.
States should set the common language. A minute spent writing an essay and a minute on YouTube are different uses; shared categories let district screen-time rules restrict the second while protecting the first.
2. States should require evidence of what works, for whom, and in what context.
Leaders need evidence that a tool improves learning and works accurately for the students who will use it. The field knows far more about whether digital tools are safe than whether they are instructionally sound. Even among the widely used purpose-built ed tech tools analyzed in Instructure’s 2026 Ed tech Evidence Report, only 21 percent have published research aligned to ESSA Tiers 1–3 evidence standards.
That gap falls hardest on students with the least margin for error. One Oakland school, which has posted double-digit literacy gains over two years, stopped using an AI-enabled literacy tool after it repeatedly mis-scored multilingual learners. Emerging research suggests that speech-recognition systems, built largely on adult speech, are consistently less accurate for young children. A tool can look effective on average and still fail the students who need it most.
States are best positioned to close evidence gaps. They should set a minimum evidence standard, covering efficacy and accuracy across student groups, and reserve state technology dollars for tools that meet it and for rigorous pilots of promising ones. A common vendor reporting format would let districts compare qualifying tools side by side without each repeating the same review.
3. Districts should audit for student learning and against redundancy.
Before adding anything new, a district should audit the platforms it already pays for and cut duplicates. Each remaining tool should align with the district’s vision for instruction and, if it does, pass a test drawn from Leading Educators’ Value Add of Technology on Teaching framework: Does it expand what educators can do (Do More), improve the quality of what they already do (Do Better), or make possible something that couldn’t happen before (Do New)? A tool that does none of these uses instructional time and money without improving students’ experiences.
Denver Public Schools shows what this can accomplish. The district cut its catalog from more than 1,000 digital tools to 346 and saved about $1 million of the roughly $21 million it spent on ed tech. Its academic team, working with operations, now makes the final call on classroom technology; each core subject has two or three prioritized tools, and specialists coach teachers on them the way they coach on curriculum.
These audits belong with the district, because districts hold the contracts. Leaders should inventory their full technology stack, keep a short list of approved tools for each core subject, and put academic leaders in charge of classroom technology decisions. Schools can then decide when their staff has the skills and support to adopt something new.
4. Districts should invest in instruction first, tools second.
The movement for high-quality instructional materials offers a precedent for this rule: strong curriculum improves outcomes when teachers get sustained, curriculum-based professional learning. Digital tools follow the same pattern.
In our pilots of AI-enabled curriculum tools in New York City, we found that teachers’ content knowledge and AI literacy were the strongest predictors of whether classroom practice changed. Yet while about 60 percent of teachers now use AI, only 18 percent have received formal guidance from their school.
Districts should budget for coaching and subject-specific professional learning with every classroom tool they buy, and states can direct implementation funding toward that support.
Who sets the rules?
A coherent system, in which each level owns a different part of the work, gives teachers, parents, and students confidence that the few tools in their classrooms are there because they help students learn.
- States set common technology categories, a minimum evidence standard for state funding, and a common reporting format for vendors. They also fund rigorous pilots.
- Districts audit their tools; keep a short, approved list for each subject; budget for coaching with every tool; and give academic leaders the final say.
- Schools protect time for coaching and check whether tools improve instruction and students’ experiences.
- Teachers decide how and when students use approved tools and make decisions about student placement, progression, and access.
- Families get a clear, short list of the tools their children use and the instructional purpose of each.
To keep these rules working, we propose that tool contracts lapse unless the district can show that a tool improves classroom experiences and learning outcomes.
In the 2024–25 school year, districts reportedly accessed an average of nearly 3,000 ed tech tools. No teacher can vet that many, and no student’s learning should depend on whether the adult at the front of the room had time to try. Rules grounded in instruction ensure leaders can answer two questions before buying any tool: How will it help students learn, and what will educators need to make that happen?
