Reddit, the popular social media and discussion platform, has filed a lawsuit against AI startup Anthropic, accusing it of stealing data from the Reddit contents to train its AI models without any permission. The complaint, filed on Wednesday in San Francisco superior court, alleged unauthorized use of third party content, despite previous assurances that it would not do so.
Anthropic, backed by the tech giant Amazon and Alphabet (Google parent company), is known for developing Claude, an AI model designed to generate human-like text responses. Reddit claims that Antropic trained their AI model Claude using Reddit’s content without entering into a licence in agreement— thereby violating Reddit’s terms of service and user policies.
According to the complaint, Anthropic’s attempted to access Reddit’s content over 100,000 times. This activity persisted even after Anthropic assured credit in July of 2024 that it has blocked what’s from accessing the platform’s contents. By scraping content and using it for commercial purposes, Anthropic violated Reddit’s user policy and “enriched itself to the tune of tens of billions of dollars,” the complaint added.
In response, Anthropic has denied the allegations, stating, “We disagree with Reddit’s claims and will defend ourselves vigorously.” The company has not provided further details on the matter.
An Anthropic spokesperson said, “We disagree with Reddit’s claims and will defend ourselves vigorously,”
Reddit’s Chief Legal Officer, Ben Lee, emphasizes the importance of clear limitations on how AI companies use scraped content. He also stated, “We believe in an open internet, but AI companies need clear limitations on how they use content they scrape.” Reddit highlighted that, unlike Anthropic, companies like Google and OpenAI have entered into licensing agreements to use Reddit’s data. Reddit has quoted Claude as admitting, in response to user queries, that it was “trained on at least some Reddit data” and that it could not confirm whether any of that material had been deleted from the platform—despite Reddit’s policy that users can remove or delete their posts at any time.
By allegedly ignoring legal rules, Reddit asserts that Anthropic “enriched itself to the tune of tens of billions of dollars” by deploying AI models trained on user-generated content without paying for or licensing it.
This lawsuit is part of a growing trend where content creators and platforms are challenging AI companies over the use of their data. The outcome of this case could set a precedent for how AI companies access and use publicly available data, potentially leading to more stringent regulations and licensing requirements.
Both companies are headquartered in San Francisco, with their offices less than a ten-minute walk apart. On 22th May, the startup introduced its latest versions of Claude—named Opus 4 and Sonnet 4. Industry insiders estimate Anthropic’s annualized revenue has reached around $3 billion. It highlights how rapidly the company has scaled by using their AI Models.
As AI technologies continue to evolve, the balance between innovation and respecting intellectual property rights remains a critical issue. The tech industry and content creators alike will be closely watching the developments of this case.
The case is formally titled Reddit Inc. v. Anthropic PBC, filed in San Francisco Superior Court under case number CGC-25-524892. A hearing date has not yet been publicly announced.
