Reflections from New York’s EdTech Week 2025.
American education is under immense stress from battles over curriculum and funding, stagnating or declining student performance, and rapid technological change. At the same time, education is also experiencing the same megatrends as other sectors, namely the rapid advancement of AI and demographic shifts toward an aging workforce. Public K-12 enrollment is projected to fall 5.5% between 2022 and 2031, with California, New Mexico, and Hawaii facing declines exceeding 15%. These trends are not limited to secondary education. Higher education is in the midst of its own reckoning, and the adult education field is grasping to define what skills and support tomorrow’s workforce will need, anticipating a massive demand for reskilling and upskilling. The World Economic Forum projects that the majority of workers will require reskilling or upskilling by 2030.
Amid this instability, the education technology sector continues to grow. Passionate innovators and investors are committed to addressing America’s educational challenges with effective solutions, and stakeholders are asking thoughtful questions about what kind of interventions are actually making a difference. Each year, EdTech Week brings together a community of leaders thinking deeply about how to solve America’s educational challenges. This year’s gathering offered an opportunity to assess whether current and emerging approaches are matching the scale of education’s challenges and opportunities.
The reality of AI adoption
As AI’s promise plays out, the challenge is pushing past superficial boosterism or stubborn critique to define how the technology might actually solve real challenges for educators, students, and families. These stakeholders have legitimate concerns about how AI might undermine the essential humanity of learning, and safety concerns demand continued scrutiny. But the sector cannot ignore the technology’s rapid arrival and widespread adoption.
According to one recent survey, nearly three-quarters of high-school students reported using AI for personal use, and half reported using it for school. The speed with which leading AI models have captured users with free, all-purpose chat tools adds urgency to efforts to prepare students to be responsible users. It also poses existential questions about the edtech industry’s operating model: What space will remain for purpose-built educational tools when students can access ChatGPT, Claude, or Gemini directly? AI chat services are fundamentally changing the competitive landscape for educational technology, though the challenge of differentiation isn’t unique to education — healthcare, legal services, and other knowledge-intensive sectors face similar questions about when custom tools add value beyond general-purpose AI.
One panel featuring funders and innovators addressed this challenge head on, with panelists predicting an upcoming contraction of the edtech market. This isn’t necessarily a bad thing. The existential threats and transformative potential of AI are deeply destabilizing but may also have a focusing effect, forcing partnerships and imaginative solutions that recently seemed unattractive or unnecessary. Investors expressed growing openness to flexible models, including revenue-sharing agreements and new kinds of partnership.
Innovation at the speed of trust
“Innovation moves at the speed of trust” was a motif throughout EdTech Week, and for good reason. Too often districts tell of working with companies building solutions based on what technology can do rather than with actual classroom problems that need solving, resulting in incomplete products that promise much but deliver little. As one superintendent put it in a “Real Talk” session, “I want partners, not vendors.” With budgets constricting and stakes rising, schools must be more discerning than ever about technology investments. Increasingly, leaders are prioritizing systems-level innovation that boldly reimagines current education models, rather than isolated technology solutions. The quality of outcomes — and the evidence supporting them — matters more than the sophistication of the tool.
But measuring those outcomes may require a more nuanced approach than education research timelines have traditionally allowed. One compelling session explored how new R&D engines could accelerate validated solution deployment while still maintaining high performance and outcome standards. Edtech companies operate on weeks-long development cycles, while rigorous studies can take years. Tiered evidence frameworks can offer a middle path: Rather than waiting for gold-standard data, organizations can demonstrate progress through incremental validation — from strong theoretical foundations to pilot data to scaled implementation results.
As school districts seek to navigate the proliferation of edtech solutions, outcomes-based contracts are emerging as a powerful trust-building mechanism that addresses both the evidence gap and implementation challenge. By tying payment to actual student progress rather than product deployment, these contracts create mutual accountability between districts and providers. On a powerhouse panel about the market advantages of outcomes-based contracting, one philanthropy leader observed that traditional funding often leaves funders feeling “like they have put so much money into the work and then everything stays the same.” Outcomes-based approaches shift the dynamic entirely, from “believe our logic model” to demonstrable impact. Critically, they also incentivize implementation fidelity: The hard part isn’t building a great product, it’s ensuring it gets used as intended. When contracts and implementation plans are designed thoughtfully, they clarify expectations and strengthen partnerships from the start.
Reimagining what’s possible
What if we truly reimagine the future of school, preserving what works best about traditional approaches while leveraging new technology’s full potential? At EdTech Week, innovators and stakeholders asked: What kinds of policies and approaches will balance competing needs — primarily students’ — while still allowing for meaningful innovation to meet the immense challenges facing education today?
One session focused on designing a new kind of high school: ASU Prep’s Amy McGrath and economist Steve Levitt presented a model where students spend roughly three hours daily using adaptive AI software to master core academics, freeing the rest of the day for what makes education distinctly human. Students participate in Socratic seminars learning to disagree with humility and curiosity. They pursue passion projects on topics as varied as the history of cake frosting, the psychology of werewolves, or game theory. The model seems promising: While the ASU Prep Tempe, powered by the Levitt Lab is still too nascent to deliver significant evidence, earlier data from the Khan World School at ASU Prep has shown that increased student engagement with responsive and adaptable curriculum can significantly increase how much core content they can master in a single year.
Convenings like EdTech Week serve as important reminders that breakthroughs can come from anywhere. Alabama is positioning itself as “the Silicon Valley of edtech testing,” with Innovate Alabama convening resources across 55 rural counties and EdFarm preparing teachers and students with digital skills statewide. Exciting case studies like this provide inspiration for the work ahead, and prove that change is possible and collaboration is essential.
