Interactive Reading Apps
30.10.2026
Journal
Checked
30 October 2026
By
The Companion Desk
Reading time
2 min
Sources
5

Facts about the episode come from WIRED's report of 3 January 2025. The lessons are our analysis, not findings from tests.

What did Fable's AI summary backlash teach reading apps?

In January 2025 WIRED reported that Fable's AI-generated 2024 reading summaries, built with OpenAI's API, made offensive comments about some readers. Fable apologised and then removed the summaries and two other AI features. The lesson for AI reading companions is about scope and labels: AI that comments on readers is riskier than AI that answers about a book, and clear labelling and off switches have since become part of how new apps present AI.

Checked 30 October 2026 · The Companion Desk · Desk review

What happened?

WIRED reported on 3 January 2025 that Fable "created an AI-powered end-of-year summary feature recapping what books users read in 2024", using "OpenAI's API". Some summaries "took on an oddly combative tone", with remarks about readers' identities that users found offensive. Fable apologised on social media: "We are deeply sorry for the hurt caused by some of our Reader Summaries this week."

Before publication, Fable's head of community told WIRED it was working on changes, including "an opt-out option for people who don't want them and clearer disclosures indicating that they're AI-generated", and had "removed the part of the model that playfully roasts the reader". After publication, WIRED updated the story: Fable "had instead made the decision to immediately remove the AI-generated 2024 reading summaries, as well as two other features that used AI."

What is the difference from an AI reading companion?

Fable's feature wrote about the reader: their tastes, their choices. AI reading companions write about the book: a recap, a cast list, an answer, a character's reply. That difference matters:

  • Writing about people invites judgments about identity, which is where Fable's summaries went wrong.
  • Writing about a book has its own risks (spoilers, invented details, misreading a character), but they are checkable against the text. See How to check an AI recap.

What has changed in how apps present AI?

Looking at listings in October 2026:

  • Labels. Thus Far says its AI Book Buddies are "labelled as Book Buddies on every post and profile" and that "A post by a reader is marked as a reader and is human."
  • Off switches. Thus Far: "Turning the Book Buddies off is one switch." Kindle's Ask this Book is opened on demand rather than inserted into the reading page.
  • Scope limits. Page-locked tools such as Magic Bookmark and Kindle's Story So Far limit what the AI may use to your place in the book, which narrows what it can say.
  • Fable itself. Its current listing describes Scout, a recommendation helper, and no AI-written reader summaries; its interactive features are human chapter rooms and annotations. See the Fable profile.

What should readers take from it?

  1. Ask what the AI writes about. The book, or you?
  2. Look for labels and switches. You should be able to tell AI text from human text and turn AI off.
  3. Check the privacy label. AI features run on your data; see Reading app privacy labels.
  4. Report bad output. Most apps have a way to flag it.

These questions are in our AI reading companion checklist.

Questions

  1. Q1

    What happened with Fable's AI summaries?

    WIRED reported in January 2025 that some AI-generated 2024 reader summaries made offensive comments about users. Fable apologised and removed the summaries and two other AI features.

  2. Q2

    Does Fable still use AI?

    Its current App Store listing describes Scout for personalised book recommendations. It does not describe AI-written reader summaries or AI answers about books.

  3. Q3

    Are AI reading companions at the same risk?

    Their risks are different. They write about the book rather than about you, so the main risks are spoilers and wrong details, which you can check against the text.

  4. Q4

    How can I tell AI posts from human posts in reading apps?

    Look for labels. Thus Far, for example, says its AI Book Buddies are labelled on every post and profile, and that reader posts are marked as human.

Sources

  1. 1WIRED: A Book App Used AI to 'Roast' Its Users. It Went Anti-Woke Instead (3 January 2025)read 2026-10-30
  2. 2Fable: Track & Discuss Books, US App Store listing (incl. App Privacy)read 2026-10-30
  3. 3Thus Far: No-Spoiler Book Club, US App Store listing (incl. App Privacy and version history)read 2026-10-30
  4. 4About Amazon: How Kindle helps you stay immersed in every book you read (published 2 April 2025, updated 2 July 2026)read 2026-10-30
  5. 5Magic Bookmark website (home page, pricing and FAQ)read 2026-10-30
  1. 01FableFable is a social reading app with spoiler-free chapter rooms, its own interactive ebook reader and a 2025 AI-summary controversy. What it does, platforms, price, privacy label and limits.03 Apps
  2. 02Thus FarThus Far is a chapter-locked book club app: posts stay hidden until you finish their chapter, with labelled AI Book Buddies and a PageKeeper spoiler check. Platforms, price, privacy label, limits.03 Apps
  3. 03AI reading companion checklistA printable checklist for choosing an AI reading companion or Kindle AI feature: place-finding, spoiler lock, accuracy, formats, price, privacy, authors' rights and age rating.02 Resources
  4. 04Interactive reading timelineEvery dated event in AI reading companions on one line: Fable's AI summaries pulled, Kindle Recaps and Ask this Book, the Authors Guild statement, Magic Bookmark and Thus Far launches.02 Resources