Funding What Works Long-Term; A Practical Roadmap for Districts

Artificial intelligence just vaulted to the top of states’ ed-tech agendas. In the latest State Educational Technology Directors’ Association (SETDA) survey, 26% of state ed-tech leaders named AI as their most pressing issue, edging out cybersecurity (21%) and professional development/tech support (18%). That’s a sharp shift from last year, when cybersecurity held the top slot. Education Week

This isn’t a hype cycle blip—it’s a signal. SETDA’s report (based on 75+ responses across 47 states) says we’re entering an “implementation” phase: more guidance, more training, and more procurement decisions that will shape classrooms for years. With ESSER relief funding expired, leaders are also grappling with how to sustain what they start. Education Week+1

Why the priority shift matters

• From pilots to policy. State officials say the guidance wave has turned into a planning-and-doing moment—AI policies must translate into training, contracts, and classroom practice. Education Week
• Funds are tighter. With pandemic dollars gone, states flag sustainability as a looming problem—even as expectations for AI-enabled tools rise. setda.org
• The stakes are higher. AI touches instruction, assessment, privacy, and safety. Decisions made this school year can lock in cost structures and classroom habits for the next five.

A practical roadmap for districts (and schools)

1) Start with purpose, not products

Define the instructional job to be done: speed up feedback on writing, differentiate reading instruction, automate routine admin tasks, or extend tutoring minutes. Tie every tool to a specific, measurable workflow (minutes saved, feedback cycles shortened, small-group targeting improved).

2) Stand up a lightweight AI governance team

Name a cross-functional group (curriculum, IT/cyber, SPED/ELL, legal/compliance, teacher leaders, student voice). Give them a 60-day charter to:

• Approve a responsible use policy (RUP) aligned to state guidance.
• Select two or three high-impact use cases to pilot.
• Establish a data checklist (see #4).

Many states now have model guidance or training you can borrow and adapt. Education Week

3) Invest in professional learning first

AI is a productivity tool before it’s a pedagogy shift. Prioritize PD that:

• Shows teachers how to draft, differentiate, and adapt materials faster.
• Models effective prompts and bias checks.
• Bakes in classroom routines (student goal-setting, citation habits, transparent tool use).

States that staffed AI “point people” and ran statewide PD cohorts are seeing traction—learn from them and localize. Education Week

4) Use a “trust & safety” procurement checklist

Before adopting an AI tool, require vendors to disclose:

• Data use & retention (no training on student data; clear deletion timelines).
• Security posture (encryption, incident response, third-party audits).
• Model transparency (features, limitations, update cadence).
• Accessibility & bias testing (ELL, disability accommodations; content filters).
• Total cost of ownership (licensing, compute, PD, and support—post-ESSER).

This helps reconcile the AI priority with the still-urgent cybersecurity concerns that state leaders continue to flag. Education Week+1

5) Pilot like a scientist

Run short, well-scoped trials (6–12 weeks) with comparison groups. Capture:

• Instructional time saved
• Feedback speed/quality (e.g., writing revision cycles)
• Differentiation (small-group formation accuracy, minutes of targeted practice)
• Student outcomes (growth measures tied to the use case)

Publish results to your community. If the pilot works, scale with an explicit phase-in plan and sunset dates for overlapping tools.

6) Center equity and inclusion

Codify expectations for human in the loop decisions, plain-language disclosure to families, and opt-out paths where appropriate. Train staff to recognize and mitigate bias, especially in language-heavy tasks and automated feedback.

7) Look for state leverage

Watch for state-negotiated pricing, model policies, or statewide PD cohorts you can join. These reduce local burden and accelerate learning. (Several states have already coordinated AI pilots or statewide supports.) Education Week

What “good” looks like this year

• Teachers have clear guardrails and time-saving workflows (e.g., planning, feedback, small-group organization).
• Students experience transparent AI use—tools that explain, show sources, and support metacognition.
• Districts publish a one-page AI overview (purpose, safeguards, approved tools, contacts).
• Procurement documents include the data/safety checklist above.
• PD is continuous, job-embedded, and aligned to your curriculum—not one-and-done.

A final note on sustainability

SETDA’s warning is plain: with federal relief gone, leaders must fund what works long-term. That means picking a few durable use cases, building staff capacity, and negotiating contracts that won’t collapse when grant dollars dry up. setda.org

---

Bottom line: States have moved AI to the top of the agenda. Schools don’t need flashy pilots; they need clear goals, strong guardrails, smart procurement, and sustained PD. If we get those right, AI can give teachers back time, personalize support, and keep human judgment at the center of learning.

Sources: Education Week coverage of SETDA’s 2025 State EdTech Trends findings; SETDA press release and report overview. Education Week+1