Reddit Accuses Perplexity of Plagiarism, Escalating AI Data Rights Dispute

Reddit (RDDT) is escalating the battle over data rights in the age of AI, filing a lawsuit against Perplexity, an AI-powered search engine startup. The suit, lodged in New York federal court, alleges that Perplexity illegally scraped user-generated content from Reddit’s platform to train its AI models. This move underscores the growing tension between content creators and AI developers regarding the fair use of copyrighted material for AI training purposes.

The lawsuit extends beyond Perplexity, targeting Oxylabs, a Lithuanian data scraping service; AWMProxy, described in the suit as a “former Russian botnet”; and SerpApi, a Texas-based startup. Reddit accuses these entities of facilitating the unauthorized extraction of its proprietary data “by masking their identities, hiding their locations and disguising their web scrapers as regular people.”

Perplexity has vehemently denied the allegations, characterizing Reddit’s actions as “extortion” and an affront to the principle of an open internet. SerpApi also issued a statement challenging Reddit’s claims and vowing to vigorously defend itself in court. This legal action follows a similar suit filed by Reddit against Anthropic in June, further solidifying Reddit’s position as a key player in the fight for data rights in the AI era.

In a statement, Reddit’s Chief Legal Officer, Ben Lee, emphasized the high stakes of the current AI landscape, noting that AI companies are engaged in an “arms race for quality human content,” which has fueled an “industrial-scale ‘data laundering’ economy.” This sentiment highlights the increasingly critical role of high-quality data in training effective AI models and the lengths to which companies may go to obtain it.

The practice of data scraping, often employed to bypass technological safeguards aimed at protecting data, has become a central issue in the debate surrounding AI development. Reddit, with its vast and vibrant ecosystem of over 100,000 interest-based “subreddit” communities, represents a particularly attractive target for those seeking training data. The lawsuit claims that Reddit’s user posts have become the most frequently cited source for AI-generated responses on Perplexity, underscoring the platform’s value as a source of human-generated content.

According to the lawsuit, Reddit issued a cease-and-desist letter to Perplexity. Following this, Reddit alleges that Perplexity increased its citations to Reddit content “forty-fold,” suggesting a deliberate attempt to circumvent Reddit’s restrictions.

The value of Reddit’s data extends beyond its sheer volume. AI researchers have noted that the platform’s moderated discussions, which encompass a wide range of topics and viewpoints, contribute to AI chatbots’ ability to generate more natural and engaging responses. This makes Reddit’s data particularly valuable for training AI models designed to interact with humans.

Recognizing the strategic importance of its data, Reddit has adopted a proactive approach by establishing AI-related licensing agreements with major players such as OpenAI and Alphabet’s Google. These agreements allow these companies to access Reddit’s data in exchange for a fee, highlighting the potential for monetization of user-generated content in the AI era.

Perplexity responded to the lawsuit on Reddit itself, maintaining that it does not train AI models on Reddit content but instead summarizes and cites publicly available discussions. According to Perplexity, this practice does not require a licensing agreement.

“A year ago, after explaining this, Reddit insisted we pay anyway, despite lawfully accessing Reddit data. Bowing to strong arm tactics just isn’t how we do business,” Perplexity stated, further suggesting the lawsuit is a strategic move by Reddit in its ongoing data licensing negotiations with Google and OpenAI.

Perplexity argues that Reddit’s business model relies on leveraging public data, adding that data licensing has become an increasingly important source of revenue for the social media platform. In February, Reddit’s COO Jen Wong revealed that AI licensing deals with Google and OpenAI accounted for nearly 10% of the company’s revenue, highlighting the significance of data licensing as a revenue stream for the company.

The lawsuit highlights the complex legal and ethical questions surrounding the use of data in AI development, and it is likely to have significant implications for the future of data licensing and the relationship between content creators and AI companies. The outcome of this case could set important precedents for the industry and shape the way AI models are trained in the years to come.