Skip to content
Submit incident
AII-1002-00689
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.Supporting sources are pending verification.

Authors sue Microsoft and Anthropic over AI training on copyrighted books

Authors sued Microsoft and Anthropic for allegedly using copyrighted books without permission to train AI models. A federal judge ruled that Anthropic's use was fair use under copyright law, but found the company had improperly sourced and stored the books from piracy websites, requiring a further proceeding on damages.

25 June 2025 Β· Anthropic Β· Intellectual property and appropriation
25 June 2025
AII-1002-00689
Quick facts
Category
Intellectual property and appropriation
Country
United States
Date
25 June 2025
Drafted by
Raidu Incident Detection

What happened

Authors filed a lawsuit against Microsoft in federal court in New York alleging that the company trained its Megatron AI model on a pirated collection of 200,000 books without permission.

In a separate case, authors Andrea Bartz, Charles Graeber and Kirk Wallace Johnson sued Anthropic, claiming the company used their copyrighted works to train its Claude AI model without authorization, allegedly sourcing books largely from piracy websites.

Reported impact

The authors claim that Megatron became able to imitate their writing styles as a result of training on the pirated books.

The authors seek an injunction against the companies' infringing activities and damages of up to $150,000 per work.

Index analysis

A federal judge in California ruled in June 2025 that Anthropic's use of copyrighted books to train Claude constituted fair use under US copyright law, finding the system's use was 'distinctly transformative' in creating new original content.

However, the same judge found that Anthropic obtained the books largely from piracy websites and stored them without permission, even though they were not ultimately used for model training, requiring a separate proceeding to determine damages.

Governance controls

The companies argue that using copyrighted content to create new works is permissible and that requiring payment for all copyrighted content used would hinder industry development.

Similar lawsuits are underway against Meta, OpenAI and other AI companies over use of copyrighted material in training.

Classification

Category
Intellectual property and appropriation
Industry
Media, social platforms and marketing
Country
United States
Harm status
Actual harm
Intent
Unknown
Timing
Post-deployment
Harm types
Economic or property
Failure modes
Training data misuse
Consequences
Litigation
AI systems
Claude, Megatron

Sources and evidence

Sources not yet published

This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.

How sources are verifiedSuggest a source for this incident