Skip to content
Submit incident
AII-1000-63699
ReportedReported by fewer than three independent publishers, with no official finding.

Greater Manchester school uses AI to remove 200 library books

A Greater Manchester school used an AI tool to identify approximately 200 books as 'inappropriate' and remove them from its library, including literary classics and age-appropriate young adult fiction. The school librarian was subsequently placed under investigation for allowing the books to remain, leading to stress and her reported resignation.

25 March 2026 Β· Greater Manchester secondary school Β· Governance, accountability and labour
25 March 2026
AII-1000-636992 sources ↓
Quick facts
Category
Governance, accountability and labour
Country
United Kingdom
Source
Daily Mail Online
Date
25 March 2026
Drafted by
Raidu Incident Detection

What happened

A secondary school in Greater Manchester used an AI tool to identify 193 books it deemed 'inappropriate', with the school planning to remove approximately 200 books from its library shelves.

The AI flagged titles including George Orwell's 1984, Stephanie Meyer's Twilight series, Michelle Obama's autobiography Becoming, and Nicholas Sparks' The Notebook, citing reasons such as 'themes of torture, violence, sexual coercion', 'mature romantic themes, sexual tension', and 'racism and political themes'.

Reported impact

The school librarian was placed under a 'safeguarding' investigation for allowing the books to remain in the library, leading her to sign off sick with stress. A complaint hearing proceeded despite her illness.

The librarian reportedly resigned from her position and disclosed the incident to Index on Censorship on condition of anonymity.

Index analysis

The AI system misclassified widely-read and age-appropriate materials as unsuitable for a school library. Twilight, marketed toward teenage audiences, was flagged for content described as inappropriate, while the graphic novel adaptation of 1984 was marked as containing graphic violence despite its educational merit.

Governance controls

The school informed the librarian that although the categorization was generated using AI, it considered the classification 'broadly accurate', indicating reliance on the AI tool's output without independent review.

The school required the librarian to remove any book deemed 'not written for children', placing responsibility on library staff to implement the AI-generated list.

Classification

Category
Governance, accountability and labour
Industry
Education and training
Country
United Kingdom
Harm status
Actual harm
Intent
Unintentional
Timing
Post-deployment
Harm types
Reputational, Psychological, Public interest, Human or fundamental rights
Harmed
Workers
Failure modes
Hallucination, Content moderation failure
Consequences
Content removal
AI systems
Unnamed AI content classification tool

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

Daily Mail Online
School accused of using AI to purge library of books including 1984
25 March 2026
Metro
School librarian sick with stress after teacher used AI to ban Twilight and 198
26 March 2026