The Literacy Mandate Race: How Two Competing Federal Bills, an 18-15 Markup, and a Union-Big Tech Coalition Are Deciding What Every Student Must Learn About AI
Two federal bills, both built around the words “AI literacy,” are working the House and Senate at the same time — and they disagree about the most basic question in the debate: who pays to teach AI, and what counts as “literacy” when the technology keeps changing faster than the curriculum.
Two Bills, One Goal, Two Philosophies
In the span of two weeks in spring 2026, Congress produced two AI-literacy bills built on opposite premises. On April 28, Senators Adam Schiff (D-CA) and Mike Rounds (R-SD) introduced the Literacy in Future Technologies Artificial Intelligence Act — the LIFT AI Act (S. 4414), with a House companion (H.R. 5584) from Reps. Tom Kean Jr. (R-NJ) and Gabe Amo (D-RI) — authorizing the National Science Foundation to make merit-based grants for AI-literacy curriculum, teacher professional development, and evaluation methods at the K-12 level [1]. On May 12, Rep. Randy Fine (R-FL) introduced the K–12 AI Literacy and Readiness Act of 2026 (H.R. 8747), a short bill that creates no grants and no new programs, and instead amends four subsections of the Elementary and Secondary Education Act of 1965 (ESEA) to add AI instruction and AI-related professional development to the existing lists of allowable uses of federal Title IV funds [2]. Senator Tim Sheehy (R-MT) introduced a Senate companion (S. 5225) on August 4 [3].
The structural difference is the whole story. The LIFT AI Act treats AI literacy as a market failure that needs federal research dollars: NSF grants to universities and nonprofits to build curricula and evaluation tools, because no single district can build rigorous AI curriculum from scratch. The Fine/Sheehy bill treats AI literacy as a permission problem: districts already receive Title IV-A money — the Student Support and Academic Enrichment grants — and the bill simply clarifies that spending it on AI instruction is allowed [2], [4]. One bill creates new supply; the other unblocks existing demand. The choice between them is less about AI than about which lever of federal education policy Congress still believes in.
The Markup: An 18-15 Vote and a Bill Waiting for the Floor
H.R. 8747 moved first. The House Committee on Education and Workforce marked it up on July 21, 2026, alongside nine other bills, and ordered it reported (amended) on an 18-15 party-line vote [2], [4]. Chairman Tim Walberg (R-MI) framed the package as advancing policies that “prepare students for the future” and “modernize our workforce system” [5]. The committee’s own description of the bill is precise about what it does: it “helps schools use federal education funds to prepare students and educators to safely and effectively use artificial intelligence while allowing schools the flexibility to direct resources where they are most needed” [5].
Both sides of the aisle say the same words — “safe, effective, and responsible” — but the vote was still straight party lines. The split suggests the dividing line is the instrument, not the goal: the LIFT AI Act would put the National Science Foundation in charge of funding — and defining — AI literacy, while H.R. 8747 leaves the definition to states and districts spending money they already have [2], [6].
Meanwhile the LIFT AI Act sits in the Senate Commerce Committee, read twice and referred on the day it was introduced, with no markup scheduled [6]. Its House companion (H.R. 5584) is tracked by the Science, Space, and Technology Committee [7]. So the practical state of play as of mid-September 2026: the permission-slip bill has cleared committee — Congress.gov’s action log shows no House floor action since the July 21 vote — and the grant bill is still waiting for a committee vote, with floor time scarce in the limited legislative days left before the midterm elections [4].
The Endorsement Lists Tell the Story
The two bills’ endorsement lists reveal the coalitions forming around AI literacy. The LIFT AI Act is endorsed by the American Federation of Teachers, Google, OpenAI, Microsoft, HP, the Information Technology Industry Council (ITI), and the Software & Information Industry Association (SIIA) [1]. That is an unusual coalition — the largest teachers’ union and the three largest AI developers on one letter — and it marks the political center of gravity: everyone agrees students should learn about AI; the argument is over the pipeline that builds the curriculum.
The Fine/Sheehy bill draws support from the opposite flank. Parents Defending Education Action, the parents’-rights advocacy group, endorsed Sheehy’s Senate companion, arguing that the bill “will give schools the flexibility to direct existing resources to address a universal concern in our education system – the safe and effective use of artificial intelligence,” and declaring that “[t]his is not a partisan issue and should not be treated as such”: “We implore the House and Senate to expedite passage of this common-sense legislation and deliver it to President Trump for his signature” [8]. The tech side is represented by CCIA, whose Michael Marn said that “with AI literacy quickly becoming a core workforce skill, students should have the opportunity to learn how to use these tools with confidence so the United States can fully capture the benefits of AI” [9].
Read together, the lists map the three-way split in the AI-in-schools debate:
- Labor + Big Tech (LIFT): union and industry aligned on grant-funded, research-backed curriculum — an alliance of convenience between groups who otherwise distrust each other’s motives. Both want someone else to build and validate the curriculum: the union, because teachers should not have to write their own; the companies, because one federally curated curriculum is cheaper to work with than fifty different state ones [1].
- Parents’ rights (Fine/Sheehy): local control, no new federal program, existing money — the constituency that wants AI taught without a new Washington apparatus deciding what “responsible” means [8].
- Workforce hawks (both bills): the competitiveness case, which is bipartisan and the easiest sell in the current climate — Senator Rounds: “We want to give American students and workers the best possible opportunity to benefit from AI-driven growth” [1], [9].
The Core Logic of Each Bill
Strip away the politics and each bill has a distinct policy logic worth understanding on its own terms.
The LIFT AI Act: literacy as a public good that markets under-supply
The LIFT AI Act’s theory is that AI literacy is a classic public good: everyone benefits when students understand AI, but no single school district has the incentive or capacity to build rigorous AI curriculum from scratch — especially when the technology changes every semester. So the federal government funds the R&D layer: NSF merit-reviewed grants to institutions of higher education and nonprofits for “educational curricula, instructional material, teacher professional development, and evaluation methods for AI literacy at the K–12 level,” with seven categories of allowable activities, including hands-on learning tools and “augmenting the existing curriculum to incorporate AI literacy where appropriate” [1], [6]. The bill even defines its target: AI literacy means “having the age-appropriate knowledge and ability to use artificial intelligence effectively, to critically interpret outputs, to solve problems in an AI-enabled world, and to mitigate potential risks” [6].
That definition is the most policy-significant sentence in either bill. It is a three-part standard: use the tools, critically interpret what they produce, and mitigate risks. Not just prompt engineering — a demand for critical interpretation built into the statute. The American Federation of Teachers, endorsing the bill, made the same point in plainer language: students should be prepared “not as passive consumers, but as critical thinkers who can use and apply AI tools on their own terms” [1]. Routing the grants through NSF rather than the Department of Education is itself a signal: the awards are merit-reviewed, and the bill’s sense-of-Congress section frames an AI-literate public as a national-security asset against “strategic adversaries” who “pursue AI technology for the purposes of surveillance, weaponization, and economic competition” [6].
The Fine/Sheehy bill: literacy as a permission problem already solved by existing law
H.R. 8747’s theory is the mirror image. The federal government does not need to fund or define AI literacy, because Title IV-A of ESEA already sends money to districts for student support and academic enrichment. The bill’s operative text consists of two phrases inserted into four statutory subsections — one allowing student instruction on “the use of artificial intelligence … in a manner that is safe, effective, and responsible,” and one allowing professional development for “teachers, paraprofessionals, school librarians and media personnel, specialized instructional support personnel, and administrators” to use and teach AI safely [2]. No new programs, no new definitions, no new money. The bill “has no fiscal impact and creates no new spending programs,” as Fine’s office put it [9].
“Artificial intelligence is reshaping our economy and our kids’ futures faster than anything since the internet, yet too many schools still don’t have the authority or the resources to keep up.” — Rep. Randy Fine (R-FL) [9]
Both bills borrow their AI definition from the same source — section 5002 of the National Artificial Intelligence Initiative Act of 2020 (15 U.S.C. 9401) [2], [6] — a rare point of genuine convergence. Whatever emerges from Congress will mean the same thing by “artificial intelligence,” anchored to a stable statutory definition that predates the current debate. What differs is everything else: who builds the curriculum, who decides what counts as “responsible,” and whether the federal role is to fund or merely to permit.
The Fine bill’s staffing list is quietly significant on its own. By explicitly covering school librarians and media personnel — not just classroom teachers — it acknowledges that AI literacy in practice happens as much in the library and media center as in the classroom, and that any AI policy built only for teachers is built for the wrong century [2].
The Skeptics: What Literacy Advocates Say the Bills Miss
Not everyone is cheering. 404 Media, in a critical look at the LIFT AI Act, noted the irony that the bill would route money through the NSF, “which has endured massive funding cuts under the Trump Administration for science research,” and observed that the bill would fund more of something that “young people and teachers alike already hate in schools” [10]. That is the sharper version of a widely shared concern: AI-literacy mandates risk becoming an ed-tech procurement vehicle — a way to get more screens and more chatbots in front of students in the name of understanding them.
The AFT itself, despite endorsing the LIFT AI Act, is moving the other direction on deployment. In a May 27, 2026 address at the National Press Club, union president Randi Weingarten demanded “an immediate end to student-facing AI in elementary schools” and a ban on “social companion” chatbots for students under 16 — while calling for schools to teach about AI within strong guardrails [11]. The position is coherent but tense: literacy first, deployment second, and never the reverse. Her framing: “getting the balance right to harness the benefits of technology while mitigating the harms,” in a ten-point plan that also calls for a “gold standard” for safety and privacy in schools, an independent research consortium whose work “should not be paid for by the industries whose products are being researched,” and a “tech tax” on Big Tech’s earnings [11].
That tension — literacy advocacy coexisting with deployment skepticism — is the real dividing line inside the AI-in-schools coalition. And the states have been running the experiment at scale: 134 AI-in-education bills were introduced across 31 states in 2026, and MultiState’s session review found states converging on the same three buckets the federal bills are fighting over — student data privacy, usage boundaries and oversight, and AI literacy and graduation requirements [12].
What to Watch
Four signals will tell you which philosophy wins:
- The House floor calendar. H.R. 8747 is out of committee and waiting for floor time in a compressed pre-election session [4]. If it passes the House, the Senate HELP Committee — where Sheehy’s companion sits — becomes the arena [3].
- NSF’s capacity. The LIFT AI Act routes money through an agency that has absorbed deep funding cuts. If the grants bill moves, the first fight will be whether there is actually money behind the authorization [10].
- The definition fight. Both bills use the same 2020 AI definition, but LIFT adds its own “AI literacy” definition with a built-in critical-interpretation requirement. If a compromise bill emerges, which definition survives is the whole ballgame [6].
- The union pivot. Weingarten’s May 2026 plan shows labor moving from endorsement to restriction. If the AFT’s position hardens, the LIFT coalition — union plus industry — gets harder to hold [11].
The deepest question is the one neither bill answers. Federal law can fund curriculum, or permit its purchase, but it cannot make an AI-literate student. That happens in classrooms, in libraries, in the difference between a teacher who understands a model’s failure modes and one who does not. Both bills are upstream inputs to that work. The race between them is really a race over which theory of the federal role — fund it, or let states buy it — gets attached to the first federal AI-education statute that actually passes.
References
[1] Schiff Senate Office — [PRIMARY] “NEWS: Sens. Schiff, Rounds Introduce Bipartisan Legislation to Bolster Artificial Intelligence Literacy for Students.” April 28, 2026. U.S. Senate press release.
[2] Congress.gov — [PRIMARY] H.R. 8747, K–12 AI Literacy and Readiness Act of 2026, 119th Congress (2025-2026). Bill text as introduced, May 12, 2026.
[3] Congress.gov — [PRIMARY] S. 5225, K–12 AI Literacy and Readiness Act of 2026, 119th Congress (2025-2026). Introduced August 4, 2026; referred to Senate HELP Committee.
[4] Advance CTE — [SECONDARY] Young, Rob. “House Republicans Advance IAA Legislation as MOA Regulations Are Rescinded.” July 24, 2026. Legislative tracking analysis.
[5] House Committee on Education and Workforce — [PRIMARY] “@EdWorkforceCmte Passes 10 Bills to Strengthen Education, Support Workers, and Expand Opportunity.” July 21, 2026. Committee press release.
[6] Congress.gov — [PRIMARY] S. 4414, LIFT AI Act, 119th Congress (2025-2026). Introduced April 28, 2026; referred to Senate Commerce Committee. CRS summary and bill text.
[7] House Committee on Science, Space, and Technology — [PRIMARY] “H.R. 5584, LIFT AI Act.” Bill page with summary, June 2026.
[8] Parents Defending Education Action — [ADVOCACY] Stoops, Terry. “Statement: PDE Action Supports Sen. Tim Sheehy’s Companion Bill to H.R. 8747.” (Parents’ rights advocacy organization.)
[9] Office of Rep. Randy Fine — [PRIMARY] “Congressman Fine Introduces K–12 AI Literacy and Readiness Act of 2026.” May 12, 2026. House press release with CCIA endorsement quote.
[10] 404 Media — [SECONDARY] Cole, Samantha. “OpenAI, Google, and Microsoft Back Bill to Fund ‘AI Literacy’ in Schools.” May 4, 2026.
[11] American Federation of Teachers — [ADVOCACY] “‘Devices Down, Eyes Up, Hands-On’: Weingarten Calls for Screen Bans, AI Limits, Active Learning, in Major National Education Address.” May 27, 2026. (National teachers’ union.)
[12] MultiState — [SECONDARY] Aaron, Izzy. “How States Are Regulating AI in Education this Legislative Session.” April 9, 2026. State legislative tracking.
AI automated and generated content. Human reviewed.
The Weekly AI Policy Brief
AI policy moves fast. Get the week’s developments, new trackers, and podcast episodes in one short email — free, ad-free, no spam.
Leave a Reply