Stock surges 7.35% following lawsuit announcement
Reddit (NASDAQ: RDDT) has filed a lawsuit against AI startup Anthropic, accusing it of unlawfully using Reddit’s platform and user-generated content to train its Claude language model—despite repeated warnings and the absence of any licensing agreement.
The lawsuit, filed Wednesday in California’s San Francisco Superior Court, claims Anthropic scraped Reddit more than 100,000 times since July 2024, continuing even after telling Reddit it had ceased such activity. The complaint describes this as a breach of Reddit’s user agreement and a case of “unfair and unlawful business acts.”
“Anthropic is in fact intentionally trained on the personal data of Reddit users without ever requesting their consent,” the filing states, calling the AI firm a “late-blooming” company that presents itself as a “white knight of the AI industry” while disregarding the platform’s rules.
Shares in Reddit jumped 7.35% on the day to close at US$118.21 after news of the suit broke.
Reddit alleges commercial exploitation
Reddit’s legal filing emphasises the value of its vast archive of user-generated discussion, describing the platform as “nearly 20 years of rich, human discussion on virtually every topic imaginable.” Chief Legal Officer Ben Lee said Reddit’s “humanity is uniquely valuable in a world flattened by AI,” arguing that companies like Anthropic have no right to exploit that data for profit without compensation.
“We will not tolerate profit-seeking entities like Anthropic commercially exploiting Reddit content for billions of dollars without any return for redditors or respect for their privacy,” Lee said in a statement.
Reddit has previously signed licensing agreements with OpenAI and Google, which allow those firms to access public Reddit data under specific terms protecting user privacy and consent. In contrast, the company claims Anthropic “refused to engage” in licensing negotiations and ignored multiple requests to cease scraping activity.
Anthropic, which has been backed by Amazon and Google, is the developer of Claude, a generative AI chatbot trained on internet data. In a 2021 paper, Anthropic acknowledged Reddit as a source of high-quality data. In this case, Reddit alleges that Claude was trained on such data despite repeated warnings and restrictions.
“Anthropic refuses to respect Reddit’s guardrails,” the complaint states. “It believes it is entitled to take whatever content it wants and use that content however it desires, with impunity.”
Focus on user agreement, not copyright
While other AI lawsuits—such as those brought by The New York Times and various book authors—centre on copyright infringement, Reddit’s suit focuses instead on breach of contract and violation of its terms of service.
The platform claims Anthropic ignored its robots.txt file, a web standard used to restrict bots from crawling certain sites, and breached specific clauses in Reddit’s user agreement prohibiting unauthorised scraping and commercial use.
The lawsuit seeks compensatory damages, restitution of any unjust enrichment, and a court injunction to prevent further use of Reddit content in Anthropic’s models. Reddit is also requesting a jury trial.
Industry context
Anthropic’s legal troubles come amid a broader wave of litigation against AI firms over training data. The startup has already faced lawsuits from authors and music publishers for allegedly using copyrighted material without permission. Other major AI players—including OpenAI, Meta, and Cohere—are also facing lawsuits from news publishers, authors, and content creators.
Anthropic, for its part, says it disagrees with Reddit’s claims and “will defend ourselves vigorously.”
Strategic stakes
The case also highlights Reddit’s growing role in the AI ecosystem. Since going public in March 2024, Reddit has positioned itself as a key content source for large language models, having secured a reported US$60 million per year deal with Google, and a separate partnership with OpenAI. CEO Sam Altman, who leads OpenAI, holds an 8.7% stake in Reddit and was formerly on its board.
Reddit currently has a market capitalisation of about US$22bn.
If successful, the lawsuit could set a precedent for how platforms assert control over their data and negotiate compensation in an era increasingly shaped by AI.
