Federal Guidance for Buying Classroom Technology Explained

Federal Guidance for Buying Classroom Technology Explained

Federal officials have given school districts little specific direction on how to evaluate AI purchases, and states and individual districts are now building their own review processes to fill that gap, according to Education Week reporting (Education Week, in an August 21 report). The scale of the problem shows in the numbers: districts have access to an average of 3,001 digital tools, but students and teachers use only four of them on average, according to an Instructure analysis of Canvas launch data covering more than 12.6 million K-12 users (Education Week, in an August 21 report).

That mismatch is surfacing as shrinking budgets and the expiration of pandemic-era federal funding push districts to scrutinize every purchase more closely (Stateline, earlier this month).

Anyone searching for federal guidance for buying classroom technology will not find a single step-by-step purchasing checklist waiting for them. The reporting shows that states and districts are developing their own approaches while districts continue to carry much of the burden themselves (Stateline, earlier this month).

What federal guidance for buying classroom technology does and does not require

Two federal privacy laws may apply, depending on the program, agency, records, and use involved. FERPA gives parents and eligible students rights over education records. PPRA applies to covered programs and activities run by state and local education agencies, and other recipients of Department of Education funding (studentprivacy.ed.gov). The Department's Model Terms of Service checklist helps schools evaluate how a given online service collects, uses, and transmits information before signing up; it functions as a review tool, not a federal approval form (studentprivacy.ed.gov). The linked federal resources do not create a blanket federal preapproval process for every AI tool a district considers, and districts still need to confirm which requirements apply to a particular tool and contract with their own counsel.

A 2023 Department of Education report goes further on philosophy than on process. It calls for keeping "humans in the loop" and directs school leaders to distrust broad vendor claims, asking six types of questions before adopting an AI tool (U.S. Department of Education, 2023). That is a separate, government-issued document from the five-question framework that circulates widely in edtech coverage. That framework was published by a longtime education-technology researcher in The 74 in 2024, not by any agency. In that April 2024 article, the researcher wrote that compelling evidence of generative AI's effects on educational outcomes did not yet exist (The 74, 2024).

Education Week reported that one privacy expert the outlet spoke with described expecting every district to confirm every vendor's compliance with state and federal privacy law as an unreasonable load for school systems that are already stretched thin (Education Week, in an August 21 report).

How individual districts are filling the gap

Sumner County Schools in Tennessee identified a specific problem, plateauing middle school reading scores, before testing any product against it. Superintendent Scott Langford said the district ran a five-day micro-pilot of Coursemojo, an AI-powered literacy tool, in a single sixth grade classroom before expanding to half its sixth grade classrooms (Stateline, earlier this month). "You have to identify what you need," Langford said.

Pennsylvania's education department directs schools to determine who controls data entered into an AI tool, review third-party data-sharing practices, limit collection to what's necessary, and keep a human in the loop for anything touching grading or discipline (Stateline, earlier this month). New York City Public Schools built similar standards into its own review process, requiring vendors to disclose their AI tools' capabilities and barring the use of student data to train AI models (Stateline, earlier this month).

Chicago Public Schools shows what enforcement looks like once a district writes these standards down. The district released an AI Guidebook and began blocking unapproved third-party AI products from its network (Education Week, in an August 21 report). A Chicago Public Schools spokesperson wrote to Stateline that the district "welcomes the promising innovation of artificial intelligence in teaching and learning, while safeguarding the development of students' foundational skills and critical thinking" (Stateline, earlier this month).

Questions to ask before buying edtech tools

None of the examples above amount to a mandated national process, but taken together, they point to a set of questions any district could organize its own review around. This list is a synthesis drawn from the reporting, not a published standard from any single agency:

• Instructional need: What defined problem is the tool meant to address?
• Evidence: What independent evidence exists for this product, for this specific use?
• Privacy: What data does the tool collect, share, retain, or use for model training?
• Human oversight: Who reviews the tool's outputs, and who has authority to stop its use?
• Equity: How does performance differ for English learners and students with disabilities?
• Contract continuity: What happens to district data and platform access if the vendor closes or is acquired?
• Renewal evidence: What usage and outcome data will determine whether the district keeps paying for it?

Testing for effectiveness and equity during a pilot

Sumner County's Langford said the strongest tools his district has tried are not the ones that replace teachers, but the ones that "help them see student understanding more quickly" (Stateline, earlier this month). The district has continued using Coursemojo since its pilot.

A pilot can also surface problems a vendor demonstration won't. A source identified only as Mak in Stateline's reporting said she heard from a middle school teacher piloting an AI tool who noticed the system was unfairly penalizing or misreading English learners' work (Stateline, earlier this month). The students noticed the pattern too, and the teacher turned it into a lesson on algorithmic bias and how AI systems are trained.

Deciding whether to sign a contract

Allentown School District in Pennsylvania runs a two-phase review that Superintendent Carol Birks describes as treating student safety and data security as non-negotiable, typically requiring a signed data processing agreement, a completed security questionnaire, and legal and privacy review before approval (Stateline, earlier this month).

Some districts are drawing on shared resources rather than building review criteria alone. The Southern Regional Education Board, a consortium of 16 Southern states, published an AI procurement and evaluation checklist for K-12 schools in 2025 (Stateline, earlier this month). California's Student Privacy Alliance offers a blanket vendor agreement with a searchable database of vendors that have already signed on, which can reduce the need for individual districts to negotiate privacy terms from scratch (Stateline, earlier this month).

Cost and vendor stability belong in this review too. In Los Angeles, the superintendent resigned after the district paid $3 million for an AI chatbot from a company called AllHere, which later filed for bankruptcy after its founder was charged with fraud (Stateline, earlier this month). That's one case, not a pattern, but it raises a fair question for any contract: what happens to district data and platform access if a vendor closes or is acquired? Dacia Toll, co-founder of Coursemojo, told Stateline the AI edtech market is "so overwhelming for districts and for vendors," adding that "most of them don't really do much that drives any kind of change for students," a vendor's assessment of her own market rather than independent evidence (Stateline, earlier this month).

Monitoring after the purchase

Buying a tool isn't the finish line. Mark Schneider, a nonresident senior fellow at the American Enterprise Institute, described a persistent "asymmetry between what the providers know and what the districts know," framing the real challenge as building mechanisms that help districts understand what they've bought and choose better next time (Stateline, earlier this month).

A source identified as Weeks in Stateline's reporting described what happens without that kind of ongoing review as "digital hoarding," saying, "What I think has happened is that I think we did a lot of digital hoarding" (Stateline, earlier this month). Reasonable checks to run before a renewal include whether complaints were filed, whether a tool's error rate holds up once used beyond the pilot group, and whether teachers are actually opening it.

What to check next

Procurement or technology staff can build a simple record of the questions asked about each tool, the evidence gathered, and who reviewed it, then route new AI purchases and upcoming renewals through that same record. Teachers should confirm which specific uses a tool is approved for, planning versus grading versus anything involving identifiable student work, before entering student data into it.

Parents and guardians can start with the district's technology office or privacy officer and ask which office approved a given tool, what data it collects, and where the district's approved-tools list, privacy notice, or board policy on AI is published. That conversation, not a search for a national standard, is the most reliable way to get an answer specific to a student's own school.