Reddit sues Anthropic, alleges unlawful data scraping to train chatbot
CALIFORNIA, UNITED STATES — Social media platform Reddit has filed a lawsuit against artificial intelligence company Anthropic, accusing it of illegally “scraping” millions of user comments to train its chatbot, Claude.
The complaint, lodged in California Superior Court in San Francisco, alleges Anthropic used automated bots to access Reddit’s content more than 100,000 times since July 2024, despite repeated requests to stop and a lack of user consent.
Reddit claims Anthropic’s actions not only breached its terms of service but also disregarded user privacy.
“AI companies should not be allowed to scrape information and content from people without clear limitations on how they can use that data,” Reddit’s chief legal officer Ben Lee told The New York Times.
He emphasized that licensing agreements with other AI firms, such as Google and OpenAI, enable Reddit to enforce protections like content deletion rights and privacy safeguards for its more than 100 million daily users.
Accusations of unfair competition
The lawsuit asserts that Anthropic, which is backed by Amazon and Google, has refused to sign licensing agreements similar to those held by other AI companies.
Reddit alleges this refusal has enabled Anthropic to commercially exploit Reddit’s data, potentially reaping “tens of billions of dollars” in benefits without accountability or compensation to Reddit or its users.
Reddit further claims that Anthropic’s actions strain its servers and diminish the user experience.
“By commercially licensing Claude for use in several of Amazon’s commercial offerings, Anthropic reaps significant profit from a technology borne of Reddit content,” the company stated in its filing.
Anthropic denies allegations
Anthropic has denied Reddit’s claims and stated it will defend itself vigorously. “We disagree with Reddit’s assertions and will defend ourselves vigorously,” a spokesperson told Ars Technica.
Reddit is seeking unspecified restitution, punitive damages, and a court injunction to prevent Anthropic from using Reddit content for commercial AI training.
Licensing and user protections
Reddit’s suit distinguishes itself from other recent legal actions against AI firms by focusing on breach of contract and unfair competition rather than copyright infringement.
Lee reiterated, “Licensing agreements enable us to enforce meaningful protections for our users, including the right to delete your content, user privacy protections, and preventing users from being spammed using this content.”
The outcome of this case could set important precedents for how user-generated content is treated in the age of artificial intelligence.
