Back to Home

How We Use AI

Decodable Reads — Chronicle LLC

Effective Date: September 21, 2026  ·  Last Updated: September 22, 2026

Looking for Literacy Leap? Literacy Leap has its own AI disclosure at decodablereads.com/literacy-leap/how-we-use-ai, written for families and for the libraries and schools that offer it.

Introduction

This page explains how Chronicle LLC uses artificial intelligence in Decodable Reads.

AI is a writing tool used to create practice materials. It is not a tutor, chatbot, or decision-maker. The instructor stays in charge of every session.

What AI is used for

When an instructor asks for new practice materials, we may use AI to draft:

  • Short decodable stories and passages on a topic the instructor or student chooses
  • Revisions that keep those stories within the phonics skills the student is ready to read
  • Comprehension questions about a generated story
  • Word lists, nonsense-word lists, dictation sentences, auditory and visual drills, and review materials matched to the student's current skill
  • Lesson plan activities for a student or a group

What AI is not used for

  • Students do not have their own AI chat and never talk to a model
  • AI does not decide what a student should learn. Placement comes from an adult-administered phonics assessment, and the instructor chooses the target skill
  • AI does not score the assessment, assign grades, or replace a teacher, tutor, or reading specialist
  • AI is not used for advertising, marketing to children, or building a profile of a child

Who is in the room

Decodable Reads is used by adult instructors and reading specialists who generate materials for the students they serve. Where an instructor has enabled student logins, a student signs in only to read passages the instructor generated for them — students never prompt a model themselves.

How a story stays decodable

Generated stories are written from the phonics skills and allowable words for that student, then checked and revised so the text stays within what the student is ready to decode. The instructor can generate again if a story is not a good fit.

Content safeguards

We take extra steps so generated stories stay appropriate for children learning to read with an adult:

  • Topic checks block profanity, slurs, sexual content, self-harm language, graphic violence, and similar requests before a story is generated
  • The checks also catch attempts to disguise a blocked word by spacing out or substituting characters
  • Typed topics cannot include web addresses or email addresses, and cannot contain instructions that try to override the writing instructions we send the model
  • The writing instructions tell the model to produce a simple, cheerful story and to substitute an everyday topic if the request is not appropriate
  • Adults see the story before the child reads it and can generate a new one

No automated filter is perfect. If a story still feels off, generate again or skip that topic. Contact us if something should not have appeared.

What is sent to the AI provider

Story generation is provided by Anthropic (Claude), a United States company, under a commercial API. Prompts may include:

  • The story or activity topic entered by the instructor or student
  • Target phonics skills, review skills, focus words, and allowable word lists
  • Generated story text when we ask the model to revise a passage or write quiz questions

We do not send student names, ages, grades, or usernames to Anthropic. Anything typed into an open topic box is sent exactly as written, including a name or other detail an instructor or student types.

We do not use student data or personal information to train, fine-tune, or otherwise develop AI models. Anthropic’s commercial API terms likewise do not allow customer content to be used to train their models. Anthropic automatically deletes inputs and outputs sent through that commercial API within 30 days. It may keep them longer if its automated trust and safety systems flag a request as violating Anthropic’s usage policy, in which case it may keep those inputs and outputs for up to 2 years and related safety scores for up to 7 years, or if the law requires Anthropic to keep them.

Data we store

We may store generated stories and related practice text in the product so the instructor can use them during a lesson, review progress, or request support. We do not sell this information. More detail is in our Privacy Policy.

Instructor choices

There is not a separate "AI off" mode for generation features, because personalized decodable text is how they work — a passage matched to one student's exact skills has to be written for that student. Instructors who prefer not to use AI-generated material simply do not use the generation features; the assessment, skill tracking, and lesson planning tools do not require them.

Instructors can review, correct, or delete student information at any time, as described in our Privacy Policy.

Questions

If you have questions about how AI is used in Decodable Reads, email support@decodablereads.com.

Chronicle LLC

3412 Lawn Ave, St. Louis, MO 63139

support@decodablereads.com