Yomiuri Shimbun Holdings has filed a lawsuit in Tokyo against US startup Perplexity, alleging copyright infringement for using its news articles and images without permission.

Yomiuri claims Perplexity accessed about 120,000 of its articles between February and June 2024.

The lawsuit seeks to stop Perplexity from using the content and requests over 2.1 billion yen (US$14.2 million) in damages.

A Yomiuri lawyer said this is the first such lawsuit by a major Japanese media firm involving generative AI and copyright.

Yomiuri warned that unauthorized use of journalism could negatively impact accurate reporting and undermine the foundation of democracy.

🔗 Source: NHK World

🧠 Food for thought

1️⃣ Global media companies increasingly challenge AI firms over content usage

Yomiuri’s lawsuit joins a growing wave of media companies worldwide taking legal action against AI firms for unauthorized content use.

Just last month, a coalition of major media companies filed a lawsuit against AI company Cohere for using their content to train AI models without permission1. Similarly, ANI Media filed a lawsuit against OpenAI for allegedly using its content to train AI without authorization2.

These cases reflect a fundamental tension between AI companies’ need for vast amounts of training data and publishers’ rights to control how their content is used commercially.

The timing suggests media companies are becoming more coordinated in their legal strategy, moving beyond individual complaints to systematic challenges of AI companies’ data collection practices.

2️⃣ High-stakes copyright damages signal the financial scale of AI content disputes

Yomiuri’s 2.1 billion yen ($14.2 million) damage claim demonstrates the substantial financial exposure AI companies face in copyright litigation.

This figure reflects both the volume of alleged infringement—120,000 articles over five months—and the potential commercial value publishers place on their content when used for AI training[original article].

The substantial amount Yomiuri seeks suggests publishers view AI companies’ use of their content as representing lost licensing revenue that could have been negotiated through proper agreements.