Sensational headlines make it seem like AI will either save public education (“AI will magically give teachers back hours in their day!”) or destroy it completely (“Students only use AI to cheat!” “AI will replace teachers!”), writes Maddy Sims, a senior fellow at the Center on Reinventing Public Education (CRPE), in an essay in The Hechinger Report.
These dueling narratives dominate public debate as state and district leaders scramble to write policies, field vendor pitches and decide whether to ban or embrace tools that often feel disconnected from what teachers and students experience in classrooms.
What gets lost is the fundamental question of what learning should look like in a world in which AI is everywhere. Last year, rather than debate whether AI belongs in schools, approximately 40 policymakers and sector leaders took stock of the roadblocks in an education system designed for a different era and wrestled with what it would take to move forward responsibly.
The group included educators, researchers, funders, parent advocates and technology experts and was convened by the Center on Reinventing Public Education. We agreed that there are several persistent myths derailing conversations about AI in education, and came up with shifts for combating them.
Myth 1: Saving time for teachers is AI’s biggest value
Many AI tools promise relief through faster lesson planning, automated grading or instant feedback. Forum participants were clear that focusing too narrowly on time savings risks locking schools more tightly into systems that were never designed to prepare students for the world they are graduating into.
The deeper issue isn’t how to use AI to save time. It’s how to create a shared vision for what high-quality, future-ready learning should look like.
- The shift: Stop asking what AI can automate. Start asking what kinds of learning experiences students deserve, and how AI might help make those possible.
Myth 2: Getting the right AI tools into classrooms is the biggest challenge
Forum participants pushed back on the idea that better tools alone will solve the problem of teachers stitching together core curricula, supplemental programs, tutoring services and AI tools with little guidance. The real challenge is to align how learning is designed and experienced in schools — and the policies meant to support that work — with the skills students need to thrive in an AI-shaped world. An app is not a learning model. A collection of tools does not add up to a strategy.
- The shift: Define coherent learning models first. Evaluate AI tools based on whether they reinforce shared goals and integrate with one another to support consistent teaching and learning practices — not whether they are novel or efficient on their own.
Myth 3: Leaders must either fix today’s schools or invent new models
Should scarce state, local and philanthropic resources be used to improve existing schools or to build entirely new models of learning?
Some participants worried that using AI to personalize lessons or improve tutoring simply props up systems that no longer work. Others emphasized the moral urgency of improving conditions for students in classrooms right now.
Participants rejected this false choice. They argued for an “ambidextrous” approach: improving teaching and learning in the present while intentionally laying the groundwork for fundamentally different models in the future.
- The shift: Leaders must ensure they do not lose sight of today’s students or of tomorrow’s possibilities. Wherever possible, near-term pilot programs should help build knowledge about broader redesign.
Myth 4: AI strategy is mostly a technical or regulatory challenge
Many states and districts have focused AI efforts on acceptable-use policies. Creating guardrails certainly matters, but when compliance eclipses learning and redesign it creates a chilling effect and educators don’t feel safe to experiment.
- The shift: Policy should build flexibility for learning and iteration in service of new models, not just act as a brake pedal to combat bad behavior.
Myth 5: The human core of education is threatened by AI
Perhaps the most powerful reframing the group came up with: The real risk isn’t that AI will replace human relationships in schools. It’s that education will fail to define and protect what is most human.
Participants consistently emphasized belonging, purpose, creativity, critical thinking and connection as essential outcomes in an AI-shaped world.
But they will be fostered only if human-centered design is intentional, not assumed.
- The shift: If AI use doesn’t support students’ connections between their learning, their lives and their futures, it won’t be transformative, no matter how advanced the technology.
The group came away with a shared recognition that efficiency won’t be enough, tools alone won’t save us and fear won’t guide the field.
The question is no longer whether AI will shape education. It is whether educators, communities and policymakers will look past the headlines and seize this moment to shape AI’s role in ways that truly serve students now and in the future.
The Hechinger Report
