AI tools are transforming classrooms—but the schools seeing real impact share one thing in common: they invest in teachers, not just technology. Educational technology promises to move teaching and learning beyond outdated, “one-size-fits-all” approaches, enabling personalized, competency-based learning that accounts for the unique variability of every student. But access to technology alone doesn’t guarantee impact. Every school has its own culture, resources, and level of teacher readiness, which means successful implementation is far more context-dependent than plug-and-play.
When Coursemojo partnered with Digital Promise to understand how its AI writing tool could be implemented with high fidelity at scale, educators became co-designers of the process from the start. Recognizing that lasting digital transformation requires deep professional learning and teacher agency—not top-down mandates—Coursemojo and Digital Promise built a Research-Practice-Industry Partnership (RPIP) to identify the coaching structures, local conditions, and support systems schools need to make AI implementation work.
Distinct from traditional product testing, RPIPs help surface educators’ underlying needs and the realities of their teaching contexts. This partnership brought together Coursemojo’s edtech developers with middle school English Language Arts educators across the country. Over a school year, 16 instructional coaches and 41 teachers piloted and co-designed the tool together with developers across more than three dozen classrooms in 19 schools and three distinct sites.
Fidelity to the tool varied across sites–some teachers used Coursemojo as intended, others barely used it at all. So what conditions help educators feel ready to bring AI into their classrooms? And how can schools sustain that momentum? Through interviews, focus groups, activity logs, and co-design sessions, one lesson stood out: implementation succeeds when schools invest in teachers, not just technology.
Even the most ambitious implementation plans stall without dedicated time. Across every site, educators named the same need: protected time to reflect on and refine how they used Coursemojo. But time alone wasn’t enough–teachers also needed to learn alongside the colleagues who support their same students. Special education teachers, English language learning specialists, and instructional leaders should be brought in early as essential partners in planning and sustaining implementation. As one coach shared: “Everybody who functions in a sixth-grade ELA classroom should have dedicated PD on Coursemojo with each other.”
Educators echoed this again and again–they learn best by watching each other use the tool, not just by watching a demo from the edtech team.
Teachers are far more likely to sustain AI implementation when it complements, rather than competes with, their school’s current priorities. Schools already committed to technological innovation, personalized learning, and AI literacy had an easier time launching Coursemojo. But coupled with professional development expectations and district-level accountability pressure, educators felt stretched thin, “pulled in many different directions,” as one instructional coach described it. Another coach put it plainly: “Every department had something different that they were learning all at the same time,” leaving little capacity for any one initiative to take root. Rather than adding one more thing to teachers’ plates, schools and edtech partners should explicitly connect AI implementation to the priorities and instructional goals teachers are already working toward.
Most edtech implementations train teachers to use specific tools. With rapidly-evolving technology tools like AI, teachers’ AI literacy means more than knowing how to use it, but also understanding how it works and evaluating its outputs. This shifts teachers from simply operating a tool to exercising instructional judgement about how to use it in service of student learning. For example, while using the feedback function of Coursemojo, educators found that the feedback function worked best with a human in the loop: students needed teacher scaffolding to fully interpret and apply the AI-generated feedback. That reinforced a broader lesson–integrating technology into teaching isn’t just about the tool itself, but about pedagogy: the underlying decisions and moves that shape how learning experiences are designed. Real implementation success requires more than tool training. Teachers need to see instructional strategy and technology working together to support student learning.
Our insights highlight a consistent theme: implementation succeeds when it is grounded in both evidence-based design and learner-centered design, and when it makes room for the realities of authentic classroom practice. Educators told us clearly–AI implementation demands contextual relevance and space to iterate. They want to see it work first: evidence of successful implementation from trusted peers in similar contexts, so they can imagine what is possible in their own classrooms.
If your school is piloting an AI tool this year, these findings point to a clear starting place: protect time for teacher collaboration, connect the tool to priorities your staff already care about, and build instructional judgement alongside technical skill.
