Media coverage can sometimes give the impression that every district is racing to become the next ‘Alpha School’, fully AI-powered and reinventing instruction overnight. But conversations with real public school leaders tell a different story. For most, progress is steady and strategic: adopting an AI tool, piloting a few trainings, and learning as they go. In fact, districts taking even these initial steps may already be among the top third of adopters. Most are prioritizing understanding over acceleration, balancing innovation with practicality, and ensuring teachers feel supported before expanding further. 

Setting the stage: Why we asked districts about AI learning 

Artificial intelligence has dominated edtech conversations for the past few years, often surrounded by uncertainty, curiosity, and plenty of debate. While headlines tend to spotlight futuristic examples, many district leaders are quietly asking more practical questions: 

• What does responsible use look like?
• How should we prepare teachers?
• What do students really need to know?  

At Frontline Education, we’re in the business of equipping K-12 administrators with tools and insights that make their work simpler, so they can focus on what really matters: giving students a path to long-term purpose. Each year, for our K-12 Lens Report, we survey administrators nationwide to take the pulse on emerging priorities and persistent pressures. 

We also check in informally throughout the year with quick polls on timely topics to understand what’s top of mind for educators. AI continues to surface as one of the most consistently charged conversations in education.  

For the past two years, we’ve asked school administrators a simple but revealing question: 

“Have you introduced AI professional learning or support in your district?”  

Here’s what they told us: 

District Adoption of AI Professional Learning: 2024 vs. 2025

While implementation of AI-related professional development nearly doubled in one year, from 17% to 36%, the reality is that a full one-third of districts still haven’t taken concrete steps toward professional learning on AI. The tension between momentum and hesitation is exactly what we wanted to explore further. 

Why so many still hesitate

As Smith explains, hesitation persists for several understandable reasons, including complexity, confusion, and legitimate concern, not just resistance to innovation.

“There’s a certain segment of the teaching staff that is adamantly against using AI by students. Mostly English and social studies teachers, although math teachers are in there, too, because most of the large language models are just bad at math. A lot of the humanities folks believe that shifting thinking over to AI will raise a generation of Americans who don’t know how to think.”

He points to four friction points every district leader can relate to:

1. Staff skepticism. Teachers fear losing the human craft of thinking and writing. In some cases, that fear shows up as resistance to student use of AI, especially in subjects where original writing and analysis are central to instruction.
2. Choice overload. Smith explains, “Because there are so many AI options out there, it’s just too diffuse for folks to get their hands on.” With so many tools, it’s hard to know where to start, or which are safe for students. Teachers need clearer guidance on what’s approved, what’s effective, and how to integrate these tools without overwhelming their workflow.
3. Outdated training. Pre-recorded modules can’t keep up with AI’s pace of change, leaving many teachers unsure how to adapt new tools to real classroom needs.
4. Real-world concerns. Not every teacher is captivated by AI. Some approach it with caution, raising ethical questions about its purpose and the environmental cost of large-scale data systems. Energy use and sustainability frequently surface in faculty conversations. Smith explains, “Some staff members have legitimate concerns of contributing to the environmental cost of AI development with data centers and energy consumption.”

Stonington’s approach: Guardrails first, experimentation next

“At the District Office, we’re very committed to making sure all of our students have an experience with AI, not just as a classroom activity, but as part of their workforce and personal development. It’s a major facet of employment across all occupations, and we feel compelled to ensure students encounter it meaningfully in middle and high school.”

– Timothy Smith

At Stonington Public Schools, leadership took a structured approach: narrow the field, protect data, and give teachers confidence to explore. Smith said, “We made the decision to go with a single AI platform and to limit usage to that because telling staff and students that they could use whatever AI they want, whether it’s Claude, ChatGPT, Gemini…it would have been impossible to provide clear direction across so many choices.”

They chose a popular AI app designed specifically for schools, not because it’s the flashiest, but because it’s safe, compliant, and transparent. Tim explained:

“I recognize it may be a lower-power model than commercial options, but the trade-off is analytics, privacy, compliance, and control. Teachers can decide which assignments are AI-approved, meaning students can use AI tools to complete those specific tasks, and they can also determine which AI features within the application students are permitted to use to assist with their work.”

The control matters. It lets the district lead with guardrails, not restrictions, offering a practical balance between innovation and accountability. By embedding thoughtful policies into how AI is used, the district connects its broader vision, including student readiness and responsible experimentation, with day-to-day teaching decisions:

• No AI-generated IEPs or 504s. “Too much would be lost in the human touch,” Smith explained. He can also see which educators are using their AI app for this purpose and follow up with direct conversations, to remind that the district’s expectation is for IEP goals and plans to remain human-written.
• Middle and high school rollout only (for now).
• Visibility into use, not surveillance. Smith can see how often staff use AI for planning, differentiation, or assessment, but not how they prompt or what output they produce. This level of insight provides evidence of ROI, showing where teachers are engaging meaningfully with the tools, and surfaces areas where additional support or curriculum refinement may be needed. For instance, when teachers with established curricula are frequently generating new lessons or activities, it can signal that those materials need review or better alignment.

Building capacity through meaningful professional learning

“Overall, I think the staff response has been very positive, and realistic. Some teachers have said, ‘I had it do something for me, and you know what? It wasn’t any better than what I could have done myself.’ That kind of honest feedback is healthy. It shows staff are experimenting, but still thinking critically about where AI adds value and where it doesn’t.”

– Timothy Smith

Before launching the PD series, Stonington distributed a districtwide survey to gauge baseline knowledge and perceptions of AI among staff. It asked how familiar teachers were with AI, how they felt about its use in schools, and where they saw potential benefits or risks. The results helped leadership tailor their sessions to meet staff where they were.

Stonington invested in five rounds of PD, tailored to different needs and perspectives across the district. The first session covered the broader landscape: what AI is, what tools are emerging, and how teachers might use them responsibly. Four sessions, co-led with a popular AI edtech app, focused on practical classroom use, from assignment design to permission settings. A fifth session, led by Smith himself, explored prompt building and critical evaluation using Google Gemini, since Stonington is a Google-based district. Additionally, the middle school principal facilitated a hands-on session to help staff translate these concepts into day-to-day practice.

“I walked them through how to challenge the models’ thinking and get it to critique itself. That’s the kind of skill we want leaders to model for staff.”

Analytics from their AI app now shape next steps. If usage spikes around lesson creation in a subject with a stable curriculum, that’s a signal to reexamine alignment. If differentiation tools are trending, that’s evidence teachers are seeking ways to meet diverse needs.

Lessons learned: Start small, learn fast, stay human

Rolling out AI in a school district is as much about culture as it is about technology. Smith emphasized that progress depends on trust, communication, and curiosity more than on any particular platform or tool. For him, success comes from modeling openness rather than mandating compliance.

Smith’s biggest insight was that sustainable change doesn’t come from mandates. It comes from motivation.

“Mandating use for folks who don’t see value is harmful. Focus on the willing, let success travel.”

It’s a strategy that’s working. Stonington’s science teachers, who are naturally curious and data-driven, became early adopters. Their success stories are sparking interest elsewhere and change is starting to spread organically.

And through it all, Smith stays grounded:

“AI can enhance how we teach, but it will never replace the relationships that make learning meaningful.”

For district leaders ready to move from talk to traction

Start with purpose, not products. Define what you want teachers and students to learn from AI, not just what you want them to use.

Pick one tool to pilot and make sure it aligns with your goals and priorities. Maybe you don’t need to biggest, strongest, or fastest model. Maybe, like Stonington, data security, analytics, and guardrails are most important to you. Find a tool that aligns.

Make PD iterative. Train, test, reflect, repeat. If possible, use analytics to see where learning sticks. If not, check in with your staff regularly to learn what they’re using, how they’re using it, and whether it’s helping.

Empower your early adopters. Showcase real teacher examples instead of enforcing compliance.

Keep it human. The goal isn’t to automate teaching, it’s to amplify the human connection at the heart of learning.

To sum up his perspective, Smith shared a vision that balances innovation with caution. He described the district’s next phase as a measured, student-focused expansion that stays grounded in human connection and adaptability.

He also acknowledged the importance of perspective:

“I always carry a bit of a caveat,” Smith said. “I worry that this AI craze could be another dot-com bubble that eventually bursts. It’s valuable and important, but we can’t put all our eggs in one basket. At the end of the day, relational, person-to- person instruction is what works, and we can’t drift too far from that.”

And ultimately, he reinforced Stonington’s balanced stance:

“As a district, we’re not committed to any single technology or ‘all in’ on AI. Our goal is to expose students to AI as part of future readiness while continuing to prioritize human connection in every classroom. We’ll keep evaluating our needs, goals, and priorities alongside what the market offers and make informed decisions to adapt as we go.”

The bottom line

The data tells two stories: progress (more districts than ever are implementing AI PD) and hesitation (a third are still waiting on the sidelines). Change in public education is never instant. But as Stonington shows, even modest steps, grounded in values and guardrails, can create meaningful momentum.

“Change doesn’t happen through mandates. It happens when people feel safe to try, reflect, and share what works.”

– Timothy Smith