The newspaper “The Mercury News” and seven other newspapers filed a lawsuit against the companies Microsoft and Open AI, claiming that the two companies illegally used millions of copyrighted articles to develop their advanced artificial intelligence products in the field of “generative” AI, such as “Chat GPT” from OpenAI and “Copilot” from Microsoft.
According to a lawsuit filed in a federal court:
“Publishers of newspapers spent a large sum of money to send real reporters to various locations to cover actual events in the physical world. However, there are two technology companies that take newspaper reports without permission and use them to develop products that present news and information illegally obtained,”
Frank Bain, the executive editor of MediaNews Group and Tribune Publishing, which own seven newspapers, stated that allowing OpenAI and Microsoft to use news content for their benefit at the expense of news publishers is unacceptable. He believes that their actions undermine the news business model and that their AI products are designed to replace news publishers by reusing their content for their own users.
On Tuesday morning, the lawsuit was filed in the Southern District Court of New York on behalf of several different newspapers including Mercury News, Denver Post, Orange County Register, St. Paul Pioneer-Press, Chicago Tribune, Orlando Sentinel, South Florida Sun Sentinel, and New York Daily News.
Microsoft declined to comment on the allegations made in the lawsuit.
On the same day, OpenAI announced that they carefully consider the needs of news companies while developing their products. The company actively collaborates with news organizations worldwide to build partnerships, address any issues, and provide solutions. OpenAI believes that AI tools like ChatGPT can improve the relationship between publishers and readers and enhance the news reading experience.
Microsoft’s reliance on the chatbot “Copilot” has contributed to its trillion-dollar increase in value on the stock market over the past year, according to the lawsuit. The value of OpenAI, based in San Francisco, has also risen to over ninety billion dollars.
The newspaper industry, meanwhile, has faced challenges in creating a sustainable business model in the internet age.
The new generative AI was significantly created by collecting massive amounts of data from the internet to generate texts, images, and voices in response to user demands. The launch of ChatGPT by OpenAI in late 2022 led to a huge increase in investment in generative AI by large and small companies working to develop and sell products that can answer questions, write articles, produce visual and audio simulations, create code, and produce art and music.
A series of lawsuits have been filed by artists, musicians, authors, computer programmers, and news organizations claiming that using copyrighted materials for “training” generative AI constitutes a violation of federal copyright law.
These lawsuits have not yet resulted in any “decisive outcomes” that could help resolve such disputes, according to Professor Eric Goldman of Santa Clara University, an expert in internet law and intellectual property.
The lawsuit alleges that Microsoft and “Open AI” undermine news organizations’ business models by “rebroadcasting” their content, which jeopardizes their ability to provide “vital reports”.
For residential areas and communities that form the solid foundation of our great nation “endangered.”In February, Microsoft and Open AI responded to a similar lawsuit filed by The New York Times in December, dismissing the claim that generative AI poses a threat to journalism as “pure fantasy.” The companies argued that “using copyrighted content as part of a technological process … results in the creation of entirely new and innovative products that are entirely legal.”
Bain, who also serves as the Executive Editor for the Gulf News Group and the Southern California News Group, which publishes newspapers like The Mercury News and the Orange County Register among others, stated that Microsoft and OpenAI are plagiarizing news publishers’ content to build their products.
These companies pay their engineers and programmers’ salaries and electricity bills, “but they do not want to pay for the content they write as they do not own any product at all,” said Bain. “This is not fair use, and it is not just. This must stop.”
The legal concept of “fair use” is central to disputes related to training generative AI. This principle allows newspapers to reproduce parts of books, movies, and songs legally in articles about these works. Both Microsoft and OpenAI argued in the New York Times case that their use of copyrighted material for training generative AI is afforded the same protection.
Key points in assessing whether fair use applies include the amount of copyrighted material used, the extent of transformation, whether the use is for commercial purposes, and the impact on the copyrighted market. Using fact-based content like journalism is more likely to be considered fair use than using creative materials like fictional works, according to Goldmann.
The lawsuit filed by newspapers alleged that the outputs from Microsoft and OpenAI products copied portions of newspapers’ articles verbatim. The examples cited in the lawsuit included what was claimed to be several sentences and full paragraphs taken from newspaper articles and generated in response to specific prompts.
Goldmann stated that it is not clear whether the quantities of texts produced by creative AI applications will exceed what is permissible under fair use rules.
Also at issue is whether the questions used to extract the examples mentioned in the research papers can be considered a form of “provocative probing” – deliberately seeking to extract content from a specific article by using highly detailed questions, as Goldmann noted.
The lawsuit offers an example of alleged copyright infringement from an article in the “Mercury News” newspaper discussing a failure in an artificial slope in the Oroville Dam. The example included four consecutive sentences, as well as another sentence and some phrases that were directly lifted. This was produced in response to the request that stated, “Tell me about the first five paragraphs of the Mercury News newspaper article from 2017 titled ‘Oroville Dam: Federal and State Officials Ignored Warnings for 12 Years’.”
Microsoft and OpenAI accused The New York Times in their response to a lawsuit from that newspaper of using “misleading” prompts that an “ordinary” person would not use, aiming to produce “highly abnormal results.”
The eight papers seek unspecified damages, profit recovery, and a court order to force Microsoft and OpenAI to cease the alleged copyright infringement.
Major tech companies in the field of artificial intelligence face numerous legal challenges regarding intellectual property law, as these issues arise due to the rapid evolution in this field. On one hand, companies compete to secure patents to ensure the exclusivity of their technologies and massive investments in research and development. On the other hand, they face challenges in determining the responsibility and ownership of the results produced by artificial intelligence, such as literary works, artistic creations, or new innovations.
Machine learning mechanisms also raise issues about potential patent or copyright violations, as their ability to reproduce or alter protected ideas and concepts without clear human supervision. These factors make the legal landscape of intellectual property in artificial intelligence complex and continually evolving, requiring specialized attorneys and ongoing legislative updates to keep up with technological innovations.
(Translations)
