Free AI Curriculum Has a Byline. Read It Before Your Department Votes.

AI companies are offering to train your faculty for free. That is not generosity. It is the business model.

The free AI curriculum usually arrives as a kindness. A vendor has built a module on responsible AI use. It maps to your general education outcomes. It comes with slides, a facilitator guide, and a certificate students can put on a resume. There is no invoice. Somebody in your provost’s office forwards it to the curriculum committee with the subject line “this looks useful.”

It probably is useful. That is what makes this hard. The material is often better than what a stretched department could build in a semester. The problem is not that the curriculum is bad. The problem is that it has a byline, and almost nobody on the committee reads it.

Somebody already ran this play, and it worked

On September 10, The Verge published an interview with New York Times education technology reporter Natasha Singer about her new book, Coding Kids. Singer spent years documenting how the technology industry embedded itself in American classrooms over the past fifteen years, and the pattern she describes should be familiar to anyone who has sat through a curriculum adoption meeting.

Her account: Apple and Microsoft each built their own curricula for Advanced Placement Computer Science Principles courses, featuring their own tools, Apple’s Swift and Microsoft’s Minecraft. Google distributed low cost Chromebooks until they were ubiquitous, then shipped Classroom to sit on top of them, which meant that by the time generative AI arrived Google was already the platform of record in a great many schools. Where companies did not supply curriculum directly, they funded nonprofits that did.

Singer puts the strangeness of it plainly. “What’s striking about this is you can imagine that parents would not want, say, Pfizer AP Biology, or Northrop Grumman AP Physics, or ExxonMobil Environmental Science AP.”

The pitch that carried it was not complicated. This is where the jobs are. Students who do not learn it fall behind. The companies will provide the materials at no cost. Singer’s reporting follows that promise to its destination, which is a cohort of graduates who did what they were told and arrived in a labor market that had moved. Her worry now is repetition. “I worry we have collective amnesia,” she told The Verge. “Have we learned any lessons from these past tech cycles in schools that are going to inform the rollout of AI?”

The same names are on the new paperwork

Here is where it gets specific rather than atmospheric.

One of the organizations in Singer’s history of the coding push is Code.org, the nonprofit behind the celebrity video and the Hour of Code campaign, funded by the industry whose products it taught.

Now open the AILit Framework, the draft AI literacy framework for primary and secondary education published at ailiteracyframework.org. Page two, in the document’s own words: “Empowering Learners for the Age of AI: An AI Literacy Framework for Primary and Secondary Education (AILit Framework) is a joint initiative of the European Commission and the Organization for Economic Cooperation and Development (OECD). Code.org and leading international experts support its development.” The same page lists the European Commission, the OECD, and Code.org as hosts of the feedback events. The framework is not fringe. It feeds the PISA 2029 Media and Artificial Intelligence Literacy assessment.

That is not a scandal and this piece is not going to pretend it is one. A nonprofit with fifteen years of classroom reach is a reasonable partner for a framework that has to land in actual classrooms, and the AILit expert group is stacked with university researchers rather than vendor staff. The point is narrower and more useful than outrage: the organizations that shaped the last classroom technology cycle are named on the paperwork for this one. That is a fact you can verify in ninety seconds, and it should change how a curriculum committee reads a document rather than whether it reads it.

The pushback is real this time, which is new

The 2010s version of this story had almost no organized resistance. The 2020s version does.

On September 2, New York City announced a one year moratorium on student AI use in classrooms for the 2026 to 2027 school year, covering roughly 600,000 public school students from 2-K through eighth grade. Mayor Zohran Mamdani’s statement did not hedge: “The tech industry wants us to believe that AI-powered early education is not only inevitable, but necessary. We do not see it that way.” Companion chatbots are banned across all grades. Teachers may still use AI for lesson planning but are barred from using it to grade assignments. High school students get access in specific instances plus AI literacy classes twice a year, and a limited pilot runs in up to five classrooms using vetted tools.

Read the policy rather than the headline. It is not anti technology. It is a set of decisions about sequence, purpose, and who does the thinking, made by the institution rather than by the supplier. That is the whole distinction.

Higher education is not a school district, which is an advantage

It is tempting for faculty to read the New York City story and file it under K12. That is a mistake, and not because the age groups are comparable. It is a mistake because of who holds the pen.

A school district adopts curriculum centrally. A department does not. In higher education the AI module that shows up in a first year seminar, the responsible use unit bolted onto a research methods course, the professional development day that teaches a specific platform, all of it passes through people who have the standing to say no, ask for the byline, or rewrite the outcome. Academic freedom is not an abstraction here. It is the procedural fact that a vendor cannot adopt anything on your campus without a faculty member agreeing to teach it.

Which means the question in front of a department is not whether to use AI, and it is not whether the free curriculum is any good. It is narrower and more answerable. Whose learning outcome is this unit teaching to?

Four questions for any free AI curriculum

Take one AI related unit currently on your campus. Not the whole program. One unit.

One. Who wrote it, and what do they sell? Find the actual byline, not the logo on the cover. Funder, author, and publisher are three different lines and they are frequently three different organizations. Write all three down.

Two. What product does a student have to open to complete it? If the assignment cannot be finished without one company’s account, the unit is teaching that account. That may still be the right call. It should be a call somebody made on purpose.

Three. What survives if you remove the product? Strip the tool names out of the learning outcome and read what is left. If a durable skill remains, you have a unit with a vendor in it. If the outcome dissolves, you have a product tutorial with an outcome written on top.

Four. What is the outcome in your words? Rewrite it once, in the language your program already uses, aimed at what a student should be able to do in five years. This is the part people skip and it is the part that does the work.

Four questions, one unit, about twenty minutes. What you end up holding is not an opinion about educational technology. It is a document you can put in front of a curriculum committee.

The scale problem nobody mentions in the meeting

Here is the honest catch, and it is the reason this stays a memo instead of becoming practice.

Twenty minutes for one unit is completely reasonable. Your program has a catalog. A department running thirty courses with an AI component in a third of them is looking at a term of evenings to do properly, which means it gets done for the one course that came up in a meeting and never for the other nine. The audit is not hard. It is repetitive, and repetitive is what dies in a semester.

That gap is the entire reason the Harvest Kernel Faculty Toolkit exists. The Outline Builder, the first stage of the Course Creation Pathway, builds a course outline from your own Course Record, which is the stored description of your course, your outcomes, your policies, and your voice. It starts from what your program already says it teaches, so the outcome on the page is yours before any tool touches it. That is the difference between auditing somebody else’s curriculum one unit at a time and writing your own at the speed a term actually moves. See the Outline Builder inside the Faculty Toolkit.

If you would rather think out loud with other instructors working through the same adoption questions, the Harvest Kernel Learning Community is free and the conversation there is already running. If this is a department or institution level decision rather than a personal one, book a free call and bring the catalog.

Read the byline

Singer told The Verge she is optimistic, and her reason is worth sitting with. “Most parents want more than just getting kids to learn to use tech,” she said. “They want kids to learn how to be informed citizens in a democracy.” Swap parents for faculty and students for graduates and the sentence survives the trip to higher education intact.

The free curriculum is not the threat. The unread byline is. And the difference between a department that gets used by this cycle and one that uses it comes down to something unglamorous: somebody on the committee turned to page two and asked who wrote this, before the vote rather than after the cohort graduated.

The materials on your desk right now already have that page. Nobody has to send you anything.

Sources

• Lauren Feiner, “Schools are catching on to Big Tech’s playbook,” The Verge, September 10, 2026. theverge.com
• Jess Weatherbed, “NYC bans AI use for students until they reach high school,” The Verge, September 2, 2026. theverge.com
• OECD, “Empowering learners for the age of AI: An AI literacy framework for primary and secondary education (Review draft),” 2025. ailiteracyframework.org