Schools are catching on to Big Tech's AI playbook
New York Times education technology reporter Natasha Singer says the AI industry is pitching itself to schools with the same force, and the same tactics, that the tech industry used to build its influence in classrooms over the past 15 years. She lays out the pattern in a new book, Coding Kids, discussed in an interview on The Vergecast: computer science skills were sold to schools and students as the ticket to high-paying jobs, tech companies supplied the curriculum and tools for free or cheap, and the schools that adopted them ended up training a generation of future customers for those same companies' products.
Singer's book traces the pattern back more than a decade, to when Apple and Microsoft each built their own Advanced Placement Computer Science Principles curricula around their own tools, Apple's Swift and Microsoft's Minecraft. She argues parents would balk at a corporate-branded class in any other subject, an idea she illustrates by imagining a "Pfizer AP Biology" or an "ExxonMobil Environmental Science AP." Separately, Google turned itself into a dominant force in education by flooding schools with low-cost Chromebook laptops through the 2010s, a rollout Singer says only accelerated during the covid-19 school shutdowns, while its Classroom app became the way teachers handed out assignments and grades. Where companies stayed a step removed, they funded nonprofits: Code.org launched with a splashy video featuring Microsoft founder Bill Gates and then-Facebook founder Mark Zuckerberg urging kids to learn to code, and founder Hadi Partovi turned the attention it drew into the nationwide "Hour of Code" campaign of hour-long introductory coding sessions.
By the time many of the students who took those courses reached the job market, Singer writes, new technology had already thrown the promised payoff into question. She says some recent graduates feel let down that they were pushed to code their whole lives only to find entry-level tech jobs diminishing, that some schools are now dropping Chromebooks, and that research in multiple countries has found adding technology to classrooms often does not meaningfully improve learning. Back in the 2010s, Singer found there was little organized pushback against the industry's coding push, with one notable exception: a group of academics built a more rigorous, methodologically tested computer science curriculum aimed at drawing a more diverse set of students into the field and teaching them to question the motives of the people building the technology.
Singer says the AI industry's pitch to schools today is pulling from that same playbook, but is meeting resistance far faster than the coding push ever did. Just last week, New York City, the country's largest school system, banned elementary and middle schools from letting students use AI in the classroom for the coming school year; a day later, Los Angeles announced an even broader restriction that also covers high school students. Singer says she isn't arguing against tech in schools, but that most parents want more than children who can operate the tools: they want kids taught to be informed citizens who can question how companies, including AI companies, are steering their choices through software design, and who feel able to push back rather than be carried along by it. She describes herself as cautiously optimistic that schools are starting to work that out.
Key facts
- New York Times reporter Natasha Singer's new book, Coding Kids, traces about 15 years of tech-industry influence over K-12 curricula, including Apple- and Microsoft-branded Advanced Placement Computer Science courses built around Swift and Minecraft.
- Google became a dominant force in schools by distributing low-cost Chromebooks through the 2010s, adoption that accelerated during covid-19 school shutdowns, plus its Classroom app for assignments and grades.
- Code.org, backed by tech companies and led by founder Hadi Partovi, grew out of a launch video featuring Microsoft founder Bill Gates and then-Facebook founder Mark Zuckerberg into the nationwide "Hour of Code" campaign of hour-long coding sessions.
- Singer argues today's AI industry is pitching schools with the same playbook, but pushback has come far faster: New York City banned elementary and middle schools from letting students use AI in the classroom for the coming school year, and a day later Los Angeles announced a broader restriction that also covers high schoolers.
- Backlash to the earlier coding push is already visible: some graduates feel let down that entry-level tech jobs are diminishing, some schools are dropping Chromebooks, and research Singer cites finds classroom technology often doesn't significantly affect learning.
Why it matters
Singer's argument is that the AI industry isn't pioneering a new relationship with schools, it's reusing one that already worked: supply free or cheap tools and curriculum, get students trained on your products, and let the promise of future jobs do the persuading. The 2010s version of that playbook didn't fully deliver on the jobs promise by the time those students graduated, while it did turn many of them into practiced users, and future customers, of the companies' software. That history is the lens through which to read the current wave of AI-in-education partnerships and pilot programs.
Who it affects
K-12 students, teachers, and parents navigating curriculum decisions made largely by school districts; the districts themselves, including New York City and Los Angeles, which are now setting the terms of AI access in classrooms; and the companies on both sides of the pattern Singer describes, the 2010s coding-era players (Google, Apple, Microsoft, Code.org) and the AI companies now pitching schools using similar methods.
How to use it
There's no product here to adopt, but Singer's book Coding Kids and her Vergecast interview are the place to get the full argument and historical detail. Her practical suggestion for parents and educators is to judge AI curricula by whether they teach students to question how the software and the company behind it are designed to steer their choices, not just how to operate the tool, and to look at what New York City and Los Angeles have already restricted as a concrete reference point for how far a district can go.
How solid is it
The throughline is one reporter's interpretation, laid out in her own new book and a podcast conversation, so it should be read as an argument rather than an independently audited finding. Several supporting details are left unspecified in the piece: it doesn't name the "one expert" behind the medical-device training analogy, doesn't cite a specific study, sample size, or list of countries for the claim that classroom technology often fails to meaningfully affect learning, and gives no exact dates for the New York City or Los Angeles restrictions beyond "last week" and "a day later."
Risks and caveats
The parallel between 2010s coding programs and today's AI pitches is an analogy Singer draws, not a like-for-like study, and the two waves could differ in ways the piece doesn't examine. The unnamed sources involved, the "one expert" on medical-device training and the unnamed academics behind the alternative curriculum, can't be checked against anything else in the article. It's also not stated whether New York City's or Los Angeles' restrictions extend to teachers' own use of AI, only to students', so read "restriction" as covering student use unless a source says otherwise.
“you can imagine that parents would not want, say, Pfizer AP Biology, or Northrop Grumman AP Physics, or ExxonMobil Environmental Science AP”
— Natasha Singer, author of Coding Kids