AFT president Randi Weingarten presented a 10-point technology plan at the National Press Club on May 27, 2026, framed as "Devices down, eyes up, hands-on." The plan prohibits all screens below Grade 3 except for documented accessibility needs, all student-facing AI including digital tutors in elementary schools, and all companion chatbots for students under 16. It also proposes a "Big Tech tax" on technology companies to fund school programs. Weingarten explicitly said she is not calling for an AI ban and that AI is "here to stay," but argued that deploying student-facing AI before students reach a specific developmental threshold produces harm rather than benefit. The plan cites an AFT survey of 3,000 teachers in which 88% reported students' attention spans have shortened, with screen use identified as the primary driver. Sources: Chalkbeat, K-12 Dive, and EdWeek, all May 27, 2026.

The AFT is the largest teachers union in the country. This plan will reach district-level policy conversations within 60 days, used both by districts seeking justification for restrictions and by districts looking for a framework that draws distinctions rather than blanket bans. High school teachers are not the direct target of the elementary provisions, but the companion chatbot ban for under-16 students is directly relevant to the secondary classroom. If you work with students who use AI for tutoring, social-emotional support, or academic assistance from a product that markets emotional responsiveness, that product is the specific target of what the AFT is proposing. This is not about general-purpose AI tools; it is about tools positioned as companions, a category that is actively growing in the EdTech market.

Education Week data shows teacher AI adoption surged from 32% in 2024 to 61% in 2025, a number cited frequently as evidence that K-12 AI integration is accelerating. Stanford research tracking more than 9,000 teachers tells a different story beneath that headline: 43% of teachers who adopted AI tools were "trial users" with minimal sustained engagement, and 16% logged in once and never returned. The pattern is consistent with what researchers call "surface adoption," where registration and initial use occur without the instructional redesign necessary for tools to change practice. Sources: Education Week, February 2026; AI Magicx Blog, May 2026 (citing Stanford research).

The 61% figure is the number administrators will cite when they argue that AI is being integrated in their district. The Stanford breakdown is the number that describes whether integration is actually happening. The gap between them is where professional development dollars, time, and policy attention are being misallocated. Teachers who tried an AI tool once and returned to existing practice are not failing to adopt; they are signaling that the tool was not embedded in an instructional workflow that made returning to it the path of least resistance. That is a design problem for tool deployment, not a motivation problem for teachers. The districts that close the gap between adoption and implementation are the ones that redesign the workflow around the tool, not the ones that run another PD session on how to use it.

The National Association of Scholars and Freedom in Education published the Cather Standards on April 28, 2026, a set of model PreK-12 ELA standards named after the novelist Willa Cather. The framework argues that Common Core-era ELA standards shifted the curriculum so far toward transferable reading and writing strategies, close reading, textual evidence, claim-evidence-reasoning, that content knowledge became a casualty. The diagnosis: students graduate knowing how to annotate a passage and structure a body paragraph but without the domain knowledge, historical, literary, scientific, cultural, that gives them something original to say. The NAS framing is politically conservative, which will limit uptake in many districts. The underlying diagnosis comes from a body of research that predates the NAS's document, including the work of E.D. Hirsch, Daniel Willingham, and the knowledge-building literacy research that has driven curriculum reforms in states like Louisiana and Mississippi over the last decade. Source: Minding the Campus, April 28, 2026.

The argument the Cather Standards are making is not that ELA teachers are doing it wrong. It is that the standards framework they are implementing has deprioritized content knowledge in favor of replicable process. If that diagnosis is accurate, the implication for instruction is not to abandon close reading but to recognize that close reading without a domain knowledge base produces thin writing, the kind where students cite three sources and build a paragraph structure around them without having an argument of their own. That specific pattern is recognizable to most ELA teachers who grade student writing. The debate about what causes it is worth engaging with independently of the political valence of the people leading it.