The BBC has issued a legal threat against US-based AI firm Perplexity, alleging copyright infringement due to the verbatim reproduction of its content by Perplexity’s chatbot. This unprecedented action by one of the world’s largest news organisations highlights growing concerns over AI models using copyrighted material without permission or compensation, sparking a wider debate on web scraping and intellectual property rights.
BBC Takes Legal Action Against AI Firm
The BBC has initiated legal proceedings against Perplexity, an AI company, demanding an immediate halt to the unauthorised use of its content. The corporation is also seeking the deletion of any BBC material held by Perplexity and financial compensation for past usage. This marks the first instance of the BBC taking such a definitive stance against an AI company.
Perplexity’s Response and BBC’s Concerns
Perplexity responded to the BBC’s claims by stating, "The BBC’s claims are just one more part of the overwhelming evidence that the BBC will do anything to preserve Google’s illegal monopoly." However, the company did not elaborate on Google’s relevance to the BBC’s position. The BBC’s letter to Perplexity’s CEO, Aravind Srinivas, explicitly states that the firm’s actions constitute "copyright infringement in the UK and breach of the BBC’s terms of use."
Key Takeaways
- The BBC’s research earlier this year found that several popular AI chatbots, including Perplexity AI, inaccurately summarised news stories, including BBC content.
- The BBC highlighted that Perplexity AI’s output regarding BBC content fell short of its Editorial Guidelines for impartial and accurate news, potentially damaging the BBC’s reputation and undermining audience trust.
- The BBC also noted that despite disallowing two of Perplexity’s crawlers via its "robots.txt" file, the company "is clearly not respecting robots.txt."
Web Scraping Under Scrutiny
The rapid proliferation of generative AI models has intensified scrutiny over their reliance on existing online material, often obtained through web scraping. This practice involves bots and crawlers automatically extracting data from websites en masse. While many organisations, including the BBC, use "robots.txt" files to deter such activity, compliance remains voluntary, and some reports suggest bots do not always adhere to these directives.
Industry-Wide Concerns
The Professional Publishers Association (PPA), representing over 300 media brands, expressed deep concern that "AI platforms are currently failing to uphold UK copyright law." The PPA stated that bots are being used to "illegally scrape publishers’ content to train their models without permission or payment," a practice that directly threatens the UK’s £4.4 billion publishing industry and its 55,000 employees.
Perplexity’s "Answer Engine" Model
Perplexity describes itself as an "answer engine," aiming to provide clear, up-to-date responses by searching the web and identifying trusted sources. The company advises users to verify the accuracy of its responses, acknowledging the common issue of AI chatbots sometimes presenting false information convincingly. Perplexity also claims it does not use website content for AI model pre-training because it does not build foundation models.
