Reddit Lawsuit Accuses Perplexity, Different AI Companies, of Stealing Knowledge

Reddit filed a lawsuit towards Perplexity, together with a number of different knowledge mining firms, accusing them of stealing the social media platform’s beneficial knowledge.

The lawsuit, filed Wednesday in Manhattan federal courtroom, mentioned the businesses illegally circumvented digital guardrails to acquire knowledge used to coach AI fashions.

Perplexity’s AI instruments used Reddit feedback to generate solutions for customers, even after the corporate agreed to not scrape Reddit’s knowledge, the lawsuit mentioned.

Reddit mentioned it despatched a cease-and-desist letter to Perplexity in Might 2024 demanding it cease scraping Reddit knowledge until it made a take care of the social media firm, as Google and OpenAI had accomplished.

Perplexity mentioned it “was not utilizing Reddit content material to coach any AI fashions and that it could respect Reddit’s robots.txt,” in line with the lawsuit. Perplexity’s citations to Reddit elevated “forty-fold after Reddit informed it to cease,” the lawsuit added.

“Fairly than respect Reddit and its customers’ rights, what Perplexity has accomplished in response is solely give you more and more devious schemes to bypass Reddit’s safety methods and insurance policies,” the lawsuit says.

In response to the lawsuit, Perplexity seems to have used third-party knowledge scrapers to bypass Reddit’s digital guardrails by taking Reddit’s content material via Google’s search engine outcomes.

“In different phrases, Perplexity’s enterprise mannequin is successfully to take Reddit’s content material from Google search outcomes, feed them into a 3rd social gathering’s LLM, and name it a brand new product,” the lawsuit says. “Whereas that enterprise mannequin has one way or the other translated right into a $20 billion valuation, it has not resulted in a willingness to pay for what others (together with Google) have.”

Perplexity spokesperson Jesse Dwyer mentioned the corporate “will at all times combat vigorously for customers’ rights to freely and pretty entry public information.”

“Our strategy stays principled and accountable as we offer factual solutions with correct AI, and we is not going to tolerate threats towards openness and the general public curiosity,” Dwyer mentioned.

The opposite defendants within the lawsuit — Oxylabs UAB, AWMProxy, and SerpApi — are companies that scrape the web for knowledge after which promote the information to different synthetic intelligence firms, in line with the lawsuit.

Reddit’s lawsuit mentioned Perplexity could have used no less than a kind of companies, and that they pulled knowledge via Google outcomes of Reddit webpages.

“In a really actual sense, these Defendants are much like would-be financial institution robbers, who, figuring out they can not get into the financial institution vault, break into the armored truck carrying the money as an alternative,” Reddit’s lawsuit alleges.

A Reddit spokesperson confirmed to Enterprise Insider that the corporate has spent tens of tens of millions of {dollars} on anti-scraping methods, which the lawsuit says these firms circumvented.

Representatives for SerpApi and Oxylabs didn’t instantly reply to a request for remark by Enterprise Insider. AWMProxy, recognized within the lawsuit as a former Russian botnet, couldn’t instantly be reached for remark.

In an announcement to Enterprise Insider, Reddit’s chief authorized officer Ben Lee mentioned Oxylabs UAB, AWM Proxy, and SerpAI had been “textbook examples” of unlawful scrapers.

“Scrapers bypass technological protections to steal knowledge, then promote it to purchasers hungry for coaching materials,” he mentioned. “Reddit is a chief goal as a result of it is one of many largest and most dynamic collections of human dialog ever created.”

This story is creating and will likely be up to date.