In January, Elon Musk’s artificial intelligence startup, xAI, announced that it would use its chatbot to develop an AI tutoring system for more than a million students in El Salvador. The announcement came on the heels of similar ones from OpenAI, which is connecting students in Kazakhstan with its ChatGPT Edu services, and from Microsoft, which is similarly equipping students and teachers in the United Arab Emirates with AI-based tools and training.
While other countries are executing on national infrastructure projects for the AI era and treating it as an economic imperative, here in the United States, we can’t seem to move past a narrative of how AI makes it easier for students to cheat. Where is the enthusiasm for how AI and other emerging technologies can support our education system? Where are the creative partnerships, the research and development teams, the initiatives to get educators up to speed? Risks are inherent with any major shift in how we learn, work, and live, but it seems that in the K-12 setting that we’re focused on the wrong ones.
SHANGHAI’S APPROACH
I recently spent time in Shanghai, with an international learning community of high-ranking school system and city officials who collaborate to identify common, high-priority problems, research best practices, and then develop effective, practical solutions that can be adapted to varying cultural and political contexts.
On this particular trip, our goal was to learn about AI policy, practice, and pedagogy in China and bring it back to the U.S. What I quickly realized is that what we’re calling “AI in education” bears almost no resemblance to how it’s being implemented in Shanghai. There, it’s not merely the adoption of technology in classrooms; it’s a philosophical and systemic adoption. While we treat AI education as another curriculum topic or tool adoption challenge, they view it as critical national infrastructure, akin to their high-speed rail system. We aren’t facing a simple gap in implementation; this is a chasm in strategic thinking.
What makes Shanghai’s approach so powerful isn’t the technology itself, but how it’s woven into the educational fabric. They’ve moved beyond “AI literacy” to “AI infusion,” where artificial intelligence becomes the underlying operating system for the entire educational experience.
AN AI ASSISTANT FOR THE TEACHER
Every teacher has an AI assistant—not as a nice-to-have, but as a standard issue. These assistants handle lesson planning, grading, analytics, and professional development. The goal isn’t replacement but amplification, freeing teachers to focus on mentorship, creativity, and human connection. Moreover, every student has a digital portrait—a comprehensive profile based on continuous, multi-dimensional data collection that adapts in real time. This isn’t surveillance for control, it’s diagnostics for growth and personalizing their learning journey.
Educators receive tailored and specific feedback on teaching patterns, courses are reengineered to emphasize relationships between concepts to help students build system understanding rather than memorizing facts, and every school is part of a virtual ecosystem that extends beyond classroom walls.
As we traveled north at 340 km per hour on the fast train out of Shanghai, I pondered not just our failed rail infrastructure, where anemic Acela trains are the best-in-class offering, but also our lackluster vision and leadership in AI for education. Too many of us still can’t imagine what AI can do to elevate pedagogy by fostering greater wonder and creativity, or enrich curriculum. Or even finally help us implement personalized learning that’s aligned to learner variability. AI can help us move productive struggle to a higher level, but not if we’re fixated on how students can use it to cheat their way through a term paper.
Perhaps most instructive of all was our visit to the East China Normal University, where education psychologists and computer scientists at the Shanghai Institute of AI in Education have established an end-to-end development pipeline—from engineering to model testing to model evaluation—showcasing the power of a true R&D higher education institution.
This isn’t simply an academic think tank; it’s one of the many R&D divisions driving the entire system and engineering tangible products at scale. They solve specific, high-value problems: AI math tutors that diagnose handwritten work and identify exact logic failures; essay systems providing nuanced feedback on ancient Chinese poetry; psychological counseling bots using cognitive behavioral therapy techniques.
INDUSTRY AI USE VERSUS EDUCATION USE
From breakthroughs in clean energy to space exploration to cutting-edge biomedical innovation, researchers and developers in the U.S. are leveraging AI to radically push the limits in other industries. Yet there is a vacuum that needs to be filled when it comes to AI in the K-12 setting. A telling moment came when a student explained why she saw AI as “more like a student than a teacher,” because she and her classmates had to teach it how to understand their assignments better. This spoke to something I’ve been thinking about a lot lately—how we aren’t necessarily in the era of artificial intelligence, but the era of human intelligence.
Our students are ready for this partnership. Our systems are the bottleneck.
If we want to bring what feels like a haphazard approach to AI in education into a coherent whole, we must build coordinated leadership across all levels—activating governors and state leaders while creating nested state district implementation pairs that ensure alignment between policy and practice. Only a multi-level approach, connecting governors, state agencies, districts, and national partners, will create the alignment needed to build our educational infrastructure rather than continuing with disconnected pilot projects. We have the innovation capacity. We have the technical expertise.
The question isn’t whether we can catch up technologically, but whether we can develop the political will and strategic coherence to build an American version of this future—one that reflects our values of local control, individual liberty, and democratic participation.
Jean-Claude Brizard is president and CEO of Digital Promise.