A growing number of teachers are using AI tools, both general-purpose and publisher-specific, to generate decodable text for phonics instruction and leveled texts adjusted to a specific reading complexity, supplementing traditional reading materials with content that can be tailored on demand to a specific student's current skill level or a specific topic of interest. This practice is spreading quickly enough to draw dedicated coverage in current education press, and it raises a related, practical question worth thinking through directly: as more source material students read and respond to in writing is itself AI-generated, what does that mean for how writing tasks built on top of that material should be designed and graded.

AI-generated reading texts, when produced well, can offer genuine instructional value: precisely calibrated complexity, content tailored to a specific student's interests, and virtually unlimited practice material at exactly the right difficulty level, advantages that are genuinely hard to replicate with a fixed, finite library of traditional texts alone. For a writing assignment built on top of this kind of source material, a student's actual task, comprehending the text, extracting and analyzing evidence, constructing a response, remains genuinely their own work, even though the source text itself was AI-generated rather than authored by a human writer.

This distinction is worth making explicit, since it's genuinely different from the more commonly discussed concern about students using AI to generate their own writing directly. A student analyzing an AI-generated source text and writing their own original response is doing authentic reading comprehension and writing work, regardless of the source material's origin, in a way that's categorically different from a student having AI generate the response itself.

What this means for grading writing built on AI-generated sources

Grading a writing task built on an AI-generated source text doesn't require any fundamentally different approach than grading writing built on a traditional source: the rubric still evaluates the student's own comprehension, analysis, and written response, independent of the source material's origin. What's worth verifying, particularly for younger or less experienced students, is that an AI-generated source text is itself accurate and genuinely well-suited to the instructional purpose, since a flawed or poorly generated source text could genuinely undermine a writing task built on top of it, regardless of how well a student responds to it.

• Verify the quality and accuracy of any AI-generated source text before building a writing assignment on top of it, especially for younger or developing readers
• Grade writing built on AI-generated source material using the same rubric standards as writing built on any traditional text
• Recognize the real, distinct difference between AI-generated source material a student analyzes and AI-generated writing a student would submit as their own
• Consider AI-generated texts specifically for differentiation, calibrating reading complexity to individual student needs in ways a fixed text library can't match
• Disclose to students and families when AI-generated source material is being used, consistent with broader transparency practices around AI use in the classroom

Why this trend is likely to keep growing

As AI-generated text quality continues improving and tools become more tightly integrated with existing reading curriculum and publisher platforms, this practice of using AI-generated source material for differentiated reading instruction is likely to become considerably more common, not less, over the coming school years, making it worth having a clear, consistent approach to grading writing built on top of this kind of material now, rather than developing that approach reactively later.

Departments thinking through their own writing assignment and grading practices this year have a genuine opportunity to build clear guidance around this specific, growing use case proactively.

A distinction worth making clear from the start

As AI-generated reading material becomes a more common part of classroom instruction, keeping the distinction clear, and clearly communicated to students, between AI-generated source material a student engages with and AI-generated writing a student might submit as their own, protects the integrity of writing assessment even as the broader instructional landscape continues to change.