Content scraping: BBC threatens Perplexity with legal action

Contents

The British Broadcasting Corporation (BBC), the UK's public service broadcaster, has threatened the AI search engine Perplexity with legal action. The BBC suspects that Perplexity is using the broadcaster's online content to further develop its AI systems. Perplexity denies this.

The Financial Times quotes from a letter from the BBC to Perplexity boss Aravind Srinivas, which is available to the newspaper. In it, the BBC claims to have evidence that Perplexity is training its "standard AI model" with BBC content. To this end, Perplexity is said to carry out content scraping –, i.e. automatically downloading large parts of internet content, regardless of the wishes of the creators.

BBC makes serious allegations

Perplexity is also said to reproduce the exact wording of BBC content, for example recently published online articles, in its own text responses to Perplexity users. Although BBC links often appear as sources, the broadcaster does not seem to like this either.

In its letter, the BBC talks about a possible injunction against Perplexity if the company does not stop the activities alleged by the BBC. The deletion of all BBC material from the AI systems used by Perplexity and financial compensation for intellectual property infringements that have already occurred are also conceivable for the BBC.

Perplexity: BBC as Google henchman

Perplexity, on the other hand, considers the accusations to be "manipulative and opportunistic", the company told the Financial Times. The British broadcaster had a fundamentally wrong understanding "of technology, the internet and intellectual property rights". Perplexity sees a different intention behind the letter: The current allegations would show how far the BBC is prepared to go "to preserve Google's illegal monopoly for its own interests".

Perplexity is apparently alluding to agreements that Google has made with numerous media companies worldwide in order to compensate them for the exploitation of their content in its own news aggregators. However, Google itself now relies on AI summaries à la Perplexity – and local publishers are up in arms against this. This is because clicks on the websites of media companies are falling massively as a result.

Perplexity also argues that the US company does not operate its own AI model, but provides its users with other large LLMs such as those from Google, OpenAI or Anthropic. The only truly "proprietary" model is based on Meta's Llama AI and serves to reduce the precision of search results and negative effects such as hallucinations.

BBC study to prove problems

The BBC, on the other hand, believes that Perplexity has damaged its reputation and trustworthiness. It had already complained about this in the past.

At the beginning of the year, the broadcaster published a study to prove that the news summaries offered by Perplexity and co. often present BBC information incorrectly or in the wrong context. To this end, those responsible for the study also proactively asked the AI services to use BBC content as a source and allowed the websites to be searched via the robots.txt file. In total, they asked 100 different questions on current news topics.

Among other things, they came to the conclusion that 23 percent of Perplexity's answers were "significantly" imprecise ("significant issues"). However, to Perplexity's credit, the proportion of "significant" issues is lower than for most other AI search engines.

At the end of last year, the BBC also complained to Apple about a false report from its Apple Intelligence service. Apple subsequently deactivated the summaries. It remains to be seen how the situation with Perplexity will develop.

(nen)