Merriam-Webster’s dictionary defines the verb plagiarize as “to steal and pass off (the ideas or words of another) as one’s own: use (another’s production) without crediting the source.” And that’s exactly what its parent company, Encyclopedia Britannica, is alleging the AI company Perplexity did with its AI answers engine, according to a complaint filed Thursday in the US District Court for the Southern District of New York.
AI companies like Perplexity are no strangers to copyright infringement lawsuits. OpenAI, Midjourney and others are currently fighting these kinds of legal claims in court. (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.)
Encyclopedia Britannica’s complaint is a bit different, however. The publisher is primarily concerned with the results Perplexity’s AI answers engine spits out, rather than how the underlying tech (a large language model) was trained.
Don’t miss any of our unbiased tech content and lab-based reviews. Add CNET as a preferred Google source.
Perplexity is also somewhat unique among AI companies. Its revenue-sharing program aims to give media companies like Encyclopedia Britannica a tiny wedge of its profit pie. At the same time, the publisher argues that the very existence of this program proves the AI company knows there’s a market for human-generated content. Instead of paying for access to that copyrighted content, Perplexity allegedly takes it without permission or payment.
What’s also different in this complaint is that Encyclopedia Britannica specifically calls out the potential reputational harm it suffers from Perplexity’s trademark infringement. When you ask Perplexity a question, the response can include a link to the original sources, along with the websites’ logos. Yet AI doesn’t always get everything right; it can make stuff up in errors called hallucinations.
Encyclopedia Britannica doesn’t want its brands’ names or trademarks anywhere near the misinformation, writing in the complaint that AI-created content “confuses and deceives Perplexity users into believing (falsely) that the hallucinations are associated with, sponsored by, or approved by” the company.
This screenshot from the lawsuit shows how Perplexity and Encyclopedia Britannica define “plagiarize.”
The publisher also alleges that Perplexity’s web-scraping bot allows for the verbatim regurgitation of its content. Similar concerns exist in The New York Times’ lawsuit against ChatGPT-maker OpenAI.
Neither Encyclopedia Britannica nor Perplexity immediately responded to a request for comment.
Copyright law offers key protections to publishers and copyright creators, and it’s become one of the most controversial legal issues in AI. Typically, if a company wants to use copyright-protected material, it has to get the permission of the rights holder and license it. Tech companies are pushing for “fair use” exceptions, a legal concept that would give them permission to use protected content without negotiating with the creators.
Claude-maker Anthropic and Meta recently won copyright lawsuits under the fair use exception. Anthropic’s judge said the AI company’s use of copyrighted material was “exceedingly transformative.” But both judges cautioned that tech companies would not always win future cases.
