Report from Engadget
In Brief – The New York Times has filed a copyright infringement lawsuit against Perplexity, the self-styled Generative AI (GAI) “answer engine”. The media giant says that it sent the AI company several cease-and-desist demands to stop using its content until the two reached an agreement, but the AI company persisted in doing so. In the lawsuit, the Times accused Perplexity of infringing on its copyrights in two main ways. First, by scraping its website to train its AI models and feed content into the likes of the Claude chatbot and Comet browser to deliver real time answers to Perplexity users. Second, the Times alleged that Perplexity’s service often reproduced copies of its articles verbatim. Finally, in a related complaint, the media company accused Perplexity of damaging its brand by falsely attributing completely fabricated information, often referred to as “hallucinations”, to the newspaper. The Chicago Tribune filed a similar but separate lawsuit against Perplexity the same day.
Context – The application of copyright law to AI model training is a foundational legal and regulatory issue facing the industry. In the US, lawsuits filed by a wide variety of IP rights holders will force courts to resolve the issue. The key question is expected to be whether AI training is “fair use” under US copyright law. Two battling court opinions were released this summer illustrating the complexities. The copyright challenges to Perplexity are somewhat different. Perplexity appears to primarily be scraping websites to make their search service timely. This matters as AI chatbots and search services functionally merge. While the largest chatbot developers are fighting arguments that they must pay for content used in basic training, a number of them are making deals with publishers to include content in chatbot results. At the same time, Google increasingly provides an AI-generated “answer” response alongside traditional links. The search giant faces widespread complaints from publishers, but they are less about copyright and more related to unfair competition and charges that its AI answers reduce traffic and ad revenue. Perplexity will likely argue that its web scraping and copying is like Google’s search indexing.