BBC Threatens to Sue Perplexity, Alleging 'Verbatim' Reproduction of Its Content

The BBC is threatening to sue AI search engine Perplexity for unauthorized use of its content, alleging the artificial intelligence company generates BBC's material "verbatim."

In a letter to Perplexity CEO Aravind Srinivas, as published by The Financial Times on Friday, the BBC alleges that Perplexity's default AI model was "trained using BBC content." The BBC said it would seek an injunction unless Perplexity stopped scraping BBC content, deleted all BBC material and submitted a "a proposal for financial compensation."

The BBC declined to comment but said reporting by the FT was accurate. 

In a statement to the FT, Perplexity said the BBC's claims are "manipulative and opportunistic" and that the broadcasting giant fundamentally doesn't understand how the technology, internet or IP law work. Perplexity also alleged that the threat of litigation shows "how far the BBC is willing to go to preserve Google's illegal monopoly for its own self-interest." 

A US judge ruled last year that Google violated antitrust law to bolster its search dominance.

Since Perplexity is an online search engine built on top of a large language model, it can answer pretty much any question asked. This means that it needs good quality information to give users satisfying answers. The BBC alleges that since Perplexity generates answers built on BBC content, that lessens the need for readers to go to the BBC directly. 

There's also concern that AI companies aren't using its journalism correctly and impartially, which could damage its reputation. The BBC alleges that 17% of Perplexity search responses had major issues, and "the most common problems were factual inaccuracies, sourcing and missing context."

Perplexity didn't immediately respond to a request for comment. 

While this is the first time the BBC has gone after an AI company, it isn't the first time Perplexity has run into issues with publishers. Outlets currently suing or threatening to sue Perplexity for copyright infringement include The Wall Street Journal along with the New York Post, Forbes and The New York Times. An investigation by Wired last year alleged that Perplexity found ways to get around blocks and scrape its content. 

In the midst of these complaints, Perplexity launched a revenue sharing program with publishers last year, which includes Fortune, Time, The Texas Tribute and Der Spiegel.

Publishers are becoming highly defensive of their content, with AI companies seeing valuations sore on the backdrop of increasingly narrow margins in media. OpenAI, the creator of ChatGPT, currently has a valuation of $300 billion and Perplexity's valuation has also soared to $14 billion. Perplexity investors include SoftBank, Nvidia and Amazon and Washington Post owner Jeff Bezos. This is while journalism has struggled in the online age, with ad dollars being siphoned by Google and attention shifting towards social media apps. Since 2005, 2,900 local newspapers have closed in the US, according to a study from Northwestern University.

(Disclosure: Ziff Davis, CNET's parent company, in April filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)