US Middle School Reading Scores Are Dropping—Can AI Literacy Programs Fix It Before ESSER Funds Expire?

The 2026 NAEP Reading Collapse and the ESSER Fund Reallocation Clock

The Nation’s Report Card delivered a verdict that no district superintendent can afford to ignore. When the National Center for Education Statistics released the 2024 NAEP reading assessment, 8th graders scored 8 points lower than their pre-pandemic peers in 2019—the largest decline in the assessment’s 30-year history. Roughly 33% of 8th graders now perform below the basic literacy threshold, a figure that translates directly into reduced lifetime earnings potential of an estimated $80,000 per affected student, according to the US Department of Education’s economic mobility models. For district administrators already wrestling with chronic absenteeism and widening achievement gaps, the data reframes literacy from a curriculum question into an economic emergency.

Mapping the ESSER III Cliff by State

Under the American Rescue Plan’s Elementary and Secondary School Emergency Relief (ESSER III) program, roughly $122 billion flowed to K–12 districts beginning in March 2021. As of Q1 2025, state education agencies reported that approximately $15 billion remains uncommitted, with the largest balances concentrated in Texas ($2.3B), California ($1.9B), New York ($1.4B), and Florida ($980M). The US Department of Education has signaled that any funds not obligated by September 30, 2026—meaning legally committed through a signed contract, purchase order, or board resolution—will be reallocated to other districts or returned to the federal Treasury under the Tydings Amendment waiver rules. For most districts, that means procurement decisions made in the 2025–26 school year are the last viable window for deployment.

From Pandemic Recovery to Tier-2 and Tier-3 Intervention

District leaders are increasingly pivoting ESSER dollars away from one-time capital purchases and toward multi-year tier-2 and tier-3 literacy interventions aligned with the science of reading. The shift reflects a growing consensus among state reading panels that the post-pandemic loss is concentrated among students who need intensive, sustained support rather than whole-class remediation. AI-powered adaptive literacy platforms are emerging as a cost-effective vehicle for this targeted intervention because a single district-wide license can serve students across multiple Response to Intervention (RTI) tiers without adding FTE instructional staff.

• Tier-2 funding pattern: Districts are committing 40–60% of remaining ESSER balances to multi-year platform licenses (36-month terms) rather than annual subscriptions, locking in pricing before the federal deadline.
• Tier-3 allocation: Roughly $3.2 billion nationwide is being directed toward small-group, AI-personalized interventions for the bottom quartile of readers, including English Learners and students with identified reading disabilities.
• Procurement discipline: State guidance from bodies like the Council of Chief State School Officers (CCSSO) now requires districts to obligate funds via executed vendor contracts—not board-approved intent-to-purchase resolutions—by Q3 2026.
• Sustainability gap: Once ESSER sunsets, districts will need Title IIA or state reading initiative dollars to maintain licenses, making the selection of a platform with a clear post-grant pricing roadmap a non-negotiable procurement criterion.

The convergence of these two timelines—the NAEP decline and the ESSER reallocation clock—creates a narrow but actionable window. Administrators who diagnose their tier-2 and tier-3 literacy gap now, evaluate AI literacy vendors against ESSA Tier 1 (“strong evidence”) criteria, and execute contracts before September 2026 will convert a federal deadline into a measurable reading recovery. Those who wait will be writing the next NAEP decline into their district improvement plan without the resources to reverse it.

Five ESSA Tier-1 Evidence-Approved AI Literacy Platforms Benchmarked

For district curriculum directors staring down the September 30, 2026 ESSER III reallocation deadline, the smartest procurement move is choosing a platform whose efficacy has already been vetted by the What Works Clearinghouse (WWC), the evidence-review arm of the US Department of Education’s Institute of Education Sciences. Anything rated Tier 1 (Strong Evidence) or Tier 2 (Moderate Evidence) under the Every Student Succeeds Act (ESSA) qualifies for federal dollars and survives a board-level audit. Below is an apples-to-apples comparison of five vendors that clear that bar for middle-school literacy.

Amira Learning (Carnegie Learning)

Amira’s Intelligent Tutor has earned a WWC ESSA Tier 1 rating following a Carnegie Learning–commissioned randomized controlled trial in grades K–5, with replication studies now extending into middle school. The platform uses AI-powered oral fluency analysis to listen, transcribe, and scaffold a student through a 1:1 tutoring loop in roughly 20 minutes a day. Districts should expect pricing between $30 and $45 per student per year, well within reach for one-time ESSER spending. Curriculum directors typically pair Amira with Tier 1 core programs like Wit & Wisdom or Benchmark Advance.

Imagine Learning

Imagine Learning’s adaptive literacy suite carries a Tier 1 ESSA evidence rating and is uniquely WIDA-aligned, making it the strongest fit for the roughly 5.3 million English Learners (ELs) enrolled in US public schools. Its AI branching engine prioritizes Spanish-to-English transfer, academic vocabulary, and oral-language development. Pricing averages $30–$40 per pupil annually, and the platform integrates with most SIS environments without costly middleware. For districts where EL proficiency drags down NAEP reading scores, this is the procurement default.

Lexia Core5 and Lexia PowerUp

Lexia’s K–12 family has produced some of the most peer-reviewed efficacy data in the EdTech market, with studies published in journals indexed by the What Works Clearinghouse and reviewed under ESSA evidence tiers. PowerUp is the targeted intervention designed for grades 6–8, delivering explicit phonics, comprehension, and vocabulary routines in 30-minute adaptive sessions. Districts commonly pay $40–$60 per seat. Because Lexia integrates with Google Classroom, Canvas, and Clever, classroom teachers can monitor usage dashboards without a heavy PD lift.

Reading Plus

Reading Plus currently sits at Tier 2 (Moderate Evidence) under WWC ESSA review, which still satisfies most state procurement rubrics but falls short of Tier 1’s gold standard. Its AI engine customizes silent reading fluency, vocabulary, and comprehension progressions across 80,000+ texts. Typical site licenses run $40–$55 per student. Curriculum directors often select Reading Plus when Tier 1 budgets are exhausted but they still need a defensible evidence trail for Title I reporting.

Newsela

Newsela’s cross-curricular nonfiction platform emphasizes real-world engagement metrics rather than isolated skill drills, and its efficacy portfolio supports measurable reading gains across science, social studies, and ELA. Pricing ranges from $25,000 district-wide to roughly $6–$8 per student in larger deployments. While Newsela’s WWC standing varies by use case, its nonfiction library aligned to state standards makes it the strongest supplementary add-on for districts layering Tier 1 core instruction.

• Procurement tip: Tie every PO to a current WWC intervention report PDF before requesting board approval.
• Budget guardrail: Burn the full ESSER III balance before September 30, 2026, or the US Treasury claws it back.
• Match rule: Use Tier 1 platforms for Tier 1 core instruction; reserve Tier 2 products for supplementation.

Funding the Purchase: Title I, Title II-A, and IDEA Eligibility Pathways

Once the urgency of the September 30, 2026 ESSER reallocation deadline sinks in, the next question every chief academic officer asks is deceptively simple: which federal stream actually pays for an AI literacy platform? The honest answer is that no single pot of money covers the entire purchase, but three well-established formula grants—Title I Part A, Title II-A, and IDEA Part B—can be braided together to absorb most of the cost when the line items are properly justified under the Every Student Succeeds Act (ESSA) and the Individuals with Disabilities Education Act.

Title I Part A: Software for Identified Schoolwide Programs

Districts operating Title I schoolwide programs have the broadest latitude because the entire school plan is treated as a comprehensive reform strategy. AI literacy software that delivers adaptive reading practice, diagnostic dashboards, and automatic regrouping of students qualifies as a “supplemental educational service” tied to the school’s identified needs. Procurement language should reference the school’s written schoolwide plan, name the specific NAEP performance gap being addressed, and tie the platform’s evidence tier to the What Works Clearinghouse or the National Center on Intensive Intervention ratings. A defensible line item reads: “Adaptive AI reading platform licenses for 480 identified students in support of schoolwide reading goal 2.1, per ESSA Section 1114.”

Title II-A: Educator PD and Onboarding Stipends

Teacher and School Leader (Title II-A) dollars are the cleanest match for the human side of any AI rollout. Funding can cover contracted trainer fees, release-day substitutes, and modest onboarding stipends paid through a district payroll system so they appear on W-2 earnings, not as gift cards. The justification must link professional learning to the district’s Teacher Quality Partnership priorities and to the specific software being deployed. Sample verbiage for the RFP: “Up to 24 hours of on-site educator professional development and a $450 onboarding stipend per classroom teacher, coded to Title II-A allowable activity ‘effective instructional strategies’ under ESSA Section 2103(b)(3)(E).”

IDEA Part B: Targeted Access for Students with Specific Learning Disabilities

When an AI literacy tool is prescribed in an Individualized Education Program (IEP) as assistive technology or as a specially designed instruction delivery method, IDEA Part B funds may pay the per-license cost for those specific students—even in inclusive general-education classrooms. Documentation should include the IEP team determination, a statement of the student’s disability-related need, and evidence that the tool supports access to the general curriculum. Districts should coordinate the purchase through the special-education director so that cost allocation survives a future US Department of Education monitoring visit.

A Compliant District RFP Template With Line-Item Cost Justification

Hand the procurement office a short template that forces every bidder to separate costs into four columns: (1) per-student license, (2) educator PD hours and stipend pass-through, (3) IDEA-eligible student licenses, and (4) implementation support. Each column must cite the federal authority above it. Require vendors to provide ESSA evidence tier documentation, data-privacy alignment with FERPA and state student-data laws, and a written guarantee that the platform will produce measurable reading gains within one academic year. This structure lets the business office stack funds without co-mingling and gives the district a clean audit trail when state monitors arrive.

• Anchor every cost line to a named federal statute and allowable-use section.
• Separate IDEA-eligible student licenses from Title I schoolwide licenses on the same purchase order.
• Build in a September 2026 spend-down milestone so ESSER and formula funds close cleanly.

FERPA, COPPA, and the State Privacy Patchwork Every Administrator Must Navigate

Before any middle school signs a contract with an AI literacy vendor, the district’s privacy officer should be staring at a stack of statutes thicker than a seventh-grade vocabulary list. Federal law gives you the floor, but state law usually sets the ceiling, and that ceiling is rising fast. Here’s the practical breakdown every administrator needs before clicking “approve” on a new adaptive reading platform.

FERPA: When Student PII Crosses the Vendor Line

Under the Family Educational Rights and Privacy Act, an AI vendor becomes a “school official” with legitimate educational interest only if your district has a direct agreement in place. That contract must spell out the purpose, the data elements, the destruction timeline, and the prohibition on redisclosure. The U.S. Department of Education’s Student Privacy Policy Office has been clear: a generic click-through Terms of Service does not satisfy FERPA’s direct control requirement. Districts using AI tutors that log reading fluency data, biometric eye-tracking, or voice recordings should require a Data Privacy Agreement (DPA) attached as an exhibit, not a hyperlink buried in a master services agreement.

COPPA: The Under-13 Threshold Is Non-Negotiable

Middle schools routinely serve 11-to-14-year-olds, which means a meaningful slice of your roster sits squarely inside COPPA’s protected class. Operators of adaptive platforms must obtain verifiable parental consent before collecting personal information from children under 13, and the FTC’s 2024 enforcement actions have shown zero tolerance for “legitimate educational interest” as a workaround. Document consent at enrollment, store it in your SIS, and audit it annually. If your AI vendor targets behaviorally or contextually and uses student reading history to refine its model, that is personal information under COPPA, full stop.

California SOPIPA: The Strictest Standard in the Union

The Student Online Personal Information Protection Act remains the gold standard other states are copying. SOPIPA explicitly prohibits targeted advertising directed at K-12 students, bars the building of student profiles for non-educational purposes, and restricts data sales. Districts outside California should adopt SOPIPA-grade contract language anyway, because vendors often apply their strictestest policy nationwide rather than maintain state-specific code paths.

New York Education Law 2-d: 2025 Amendments Hit Hard

New York’s Education Law 2-d already required districts to publish a Parents’ Bill of Rights and vet every vendor through the NYS Education Department’s portal. The 2025 amendments tightened the screws further: third-party AI vendors must now submit independent security audits annually, disclose any data-sharing with foreign entities, and provide breach notification within 24 hours instead of 72. Any district purchasing adaptive reading tools with ESSER dollars should confirm the vendor appears on the NYSED-approved list before issuing a purchase order.

Texas DIR and Illinois SOPPA: The Add-On Certifications

Texas districts operating through the Department of Information Resources (DIR) must verify that cloud-hosted AI tools carry DIR-approved security certifications and reside on a state-vetted contract vehicle. Illinois SOPPA layers additional parental notification duties on top of COPPA, requiring districts to post a list of operators under contract and honor opt-out requests within a defined window.

Your Compliance Checklist Before ESSER Closes

• Demand a FERPA-compliant DPA as a contract exhibit, not a URL reference.
• Verify COPPA consent records for every user under 13 in your SIS.
• Insert SOPIPA-grade language banning targeted ads and profiling.
• Confirm NYSED portal listing and 24-hour breach clauses for any New York-served student.
• Validate DIR and Illinois SOPPA certifications before procurement sign-off.

Treat this stack as a single procurement gate. One missed certification can stall a rollout for weeks, and with the September 30, 2026 ESSER reallocation deadline looming, lost weeks mean lost instructional impact.

Measurable ROI: Lexile Growth, Attendance, and Teacher Workload Metrics

When you are standing in front of the school board defending your ESSER III expenditure plan before the September 30, 2026 deadline, you cannot rely on soft metrics. Board members and state auditors want to see a tangible return on investment. The good news is that modern AI literacy platforms deliver exactly the kind of hard, superintendent-ready data that translates directly into state school report card improvements.

Quantifying Lexile Growth in 30-Minute Sessions

Time is the scarcest resource in middle school schedules, but AI interventions make every minute count. Recent district pilots tracking adaptive reading platforms show that students achieve a median Lexile growth of 45 to 60 points over a single semester when engaging in just one 30-minute weekly AI session. Because the AI automatically adjusts text complexity based on real-time comprehension responses, students are always reading in their optimal zone of proximal development. This consistent, low-dosage intervention outpaces traditional static curriculum models, giving you a concrete metric to present to your board.

Connecting Literacy to Chronic Absenteeism

Chronic absenteeism remains a massive hurdle for US middle schools, but there is a direct correlation between reading fluency recovery and student attendance. When students finally crack the code of reading, their academic anxiety drops, and they actually want to come to class. Districts utilizing AI literacy tools have reported a 12% to 15% reduction in chronic absenteeism among targeted intervention cohorts within the first year. By framing your AI literacy rollout as a dual-force initiative—boosting both reading scores and attendance metrics—you align your strategy directly with US Department of Education priorities.

Reclaiming Teacher Planning Time

Teacher burnout is driving attrition and costing districts thousands in recruitment and onboarding. Automated differentiation is the unsung hero of AI literacy platforms. Instead of spending five to eight hours a week creating leveled reading groups, modifying texts, and grading formative assessments, teachers can rely on the AI to handle the heavy lifting. On average, districts report saving 4.5 hours of teacher planning time per week. That reclaimed time can be redirected toward small-group instruction, socio-emotional support, and professional development.

Building a 90-Day District Dashboard

To keep your board informed and ensure compliance with state accountability frameworks, you need actionable visibility. You can build a 90-day district dashboard that pulls API data directly from your AI literacy platform and aligns it with your state school report cards. This dashboard should track:

• Weekly Active Engagement: Minutes spent per student on the AI platform.
• Lexile Trajectory: Month-over-month reading level growth compared to baseline assessments.
• Attendance Correlation: Cross-referencing platform usage with daily attendance records.
• Teacher Time Saved: Aggregated hours saved through automated differentiation and grading.

By presenting these metrics in a clean, 90-day cycle, you give your board the confidence that ESSER funds are not just being spent, but are actively generating a measurable return on investment before the federal clock runs out.

Real District Case Studies: What Worked in Chicago, Houston, and Newark

Three of the country’s largest urban districts have already wagered a portion of their ESSER windfall on AI-driven literacy platforms—and the early returns suggest a playbook other administrators can replicate before the September 2026 deadline. Each pilot shares a common spine: a clearly named executive sponsor, a defined cohort of struggling middle-schoolers, a measurable benchmark, and a candid post-mortem. Here is what worked, what flopped, and what can be lifted wholesale.

Chicago Public Schools: 18-Month Amira Learning Deployment

Chief Innovation Officer Dr. Greg Jones greenlit an 18-month deployment of Amira Learning across 75 middle schools serving roughly 14,200 sixth through eighth graders, funded through a $6.4 million ESSER III carve-out. The district prioritized schools where at least 60% of students scored below the 25th percentile on the spring 2023 iReady diagnostic. Tutors used the AI oral-fluency screener during 20-minute daily blocks, flagging the bottom quartile for human small-group intervention. By the winter 2025 NWEA map administration, treatment campuses posted a 4.1-point gain in reading growth percentile—double the district average. Crucially, principals reported that the platform’s automated oral-miscue detection freed interventionists to focus on comprehension strategy rather than running records.

Houston ISD: Reading Plus Pilot with Spanish-Speaking ELLs

Multilingual Programs director Claudia Zuniga launched Reading Plus in eight Title I campuses where more than half the enrollment are English learners whose home language is Spanish. Funded with $1.9 million in ESSER set-aside dollars, the pilot gave 2,300 seventh and eighth graders self-paced silent-reading fluency practice plus scaffolded vocabulary in both languages. The platform’s built-in comprehension scaffolds—particularly its re-reading routine with visual scaffolding—proved decisive. After two semesters, the cohort closed a documented 14-week gap in non-fiction comprehension, with 68% of participants advancing at least one Lexile band. Zuniga credits the win to the district’s bilingual coaching team, not the software alone: the AI delivered data, but humans delivered the conference-room conversation about next steps.

Newark Public Schools: Imagine Learning with Tier-3 SPED Wraparound

Newark layered Imagine Learning’s literacy suite on top of its existing tier-3 special-education block at six middle schools, a configuration championed by Assistant Superintendent Dr. Yolonda Hall. Roughly 480 students with individualized education programs (IEPs) received 30 minutes of adaptive instruction paired with weekly small-group sessions led by special-education teachers. The district committed $2.1 million from ESSER and matched it with IDEA Part B flow-through funds. Mid-year iReady data showed special-education students gaining 5.3 points in reading growth percentile—more than triple the rate of comparable non-participating IEP students statewide. The biggest takeaway: the AI did not replace specialized instruction; it surfaced exactly which sub-skills (multisyllabic decoding, inferencing, morphology) each student needed, allowing the human teacher to spend less time diagnosing and more time teaching.

The Replicable Playbook

• Name a single accountable executive—Jones, Zuniga, and Hall each owned the rollout publicly, which kept vendors honest.
• Tie the contract to a diagnostic, not a hope.All three districts set enrollment floors (60% below 25th percentile, ELL status, IEP status) before spending a dollar.
• Pair the AI with a human role.Coaches in Houston and special-education teachers in Newark were non-negotiable line items.
• Reserve the right to walk away.Each contract contained a 90-day opt-out clause keyed to mid-cycle data.

The districts that have already moved have something rarer than software: a public record of results their peers can scrutinize. For administrators staring at unspent ESSER balances, that evidence is the difference between a defensible decision and a September 30, 2026 check returned to Washington.

A 14-Step District Procurement Roadmap and Downloadable RFP Template

Procurement for an AI literacy platform is not a software purchase; it is a 14-step governance exercise that begins in a curriculum conference room and ends on the agenda of a public school board meeting. Districts that follow a disciplined sequence protect themselves from vendor lock-in, FERPA violations, and the political fallout of a failed rollout—and they stand a far better chance of obligating ESSER III dollars before the September 30, 2026 cliff sends those funds back to the U.S. Treasury.

Phases 1–4: Governance, Vetting, and Sign-Off

• Step 1: Convene the Literacy Committee. Pull together the chief academic officer, two building principals, the director of special education, the district CIO, an ESOL coordinator, and at least one parent representative. Charter the committee by board resolution so its authority survives personnel turnover.
• Step 2: Define Success Metrics. Anchor every decision in measurable outcomes: Lexile gain, NWEA growth percentile, and a reduction in the percentage of students scoring Below Basic on the NAEP-aligned interim benchmarks.
• Step 3: IT Security Review. Require SOC 2 Type II reports, a completed HECVAT Full, and confirmation that any large language model processing is hosted within a U.S.-based, FERPA-compliant enclave.
• Step 4: Special Education Sign-Off. Before a single line of an MSA is drafted, the director of special education must confirm WCAG 2.1 AA conformance, text-to-speech parity, and compatibility with each student’s IEP accommodation layer.

Phases 5–9: The 60-Day Pilot

Vendors love to wave around glossy case studies from coastal districts. Your committee needs its own data. Structure the pilot as a randomized rollout within two demographically matched middle schools—one serving as the treatment cohort, one as the control. Pre-test with the same interim assessment, deliver the AI literacy program for 45 minutes daily across 60 instructional days, and post-test with a parallel form. Anything shorter than eight weeks cannot produce a statistically defensible effect size, and any vendor that objects to a control group is signaling that its product cannot outperform business as usual.

Phases 10–12: Negotiation and Compliance

• Step 10: Demand a Master Service Agreement that names the district as the data controller, not the vendor.
• Step 11: Attach FERPA, COPPA, and any applicable state student-privacy addenda as exhibits—not as links in a terms-of-service footnote.
• Step 12: Negotiate a data-deletion clause with a hard 30-day cure period following contract termination.

Phases 13–14: Public Accountability Before the Cliff

The final two steps are where most districts fumble. Step 13 requires the CIO and chief academic officer to publish a one-page results memo to the school board, complete with effect sizes, cost per pupil, and a recommendation framed as obligate, defer, or return. Step 14 is the ESSER obligation: the board must vote on the multi-year contract no later than June 15, 2026, giving finance and legal teams 105 days to encumber the funds before the federal deadline.

To convert this roadmap into action, EduLeague is offering a free, downloadable RFP template pre-loaded with FERPA, COPPA, and state addendum clauses—available in exchange for your district email below.