The New York Times Sues OpenAI and Microsoft for Using its Stories to Train Chatbots

The New York Times Sues OpenAI

The New York Times Sues OpenAI, NEW YORK (AP) — In response to the danger artificial intelligence poses to the journalism industry, The New York Times filed a federal lawsuit on Wednesday against Microsoft and OpenAI to stop the practice of utilizing its articles as training material for chatbots.

The New York Times Sues OpenAI and Microsoft for Using its Stories

The New York Times Sues OpenAIThe Times claims that the firms are endangering its ability to make a living by essentially taking billions of dollars worth of labor from its journalists. In certain instances, the companies are even spitting out Times articles verbatim to individuals using ChatGPT, generative artificial intelligence from OpenAI. The case was brought by the newspaper in Manhattan federal court.

Microsoft and OpenAI did not reply to requests for comments. The media is just one of several sectors that artificial intelligence’s explosive growth might upend.

While several publications, most notably the Times, have successfully carved out a digital area, media institutions have already been battered by the exodus of readers to online platforms. However, AI could pose a danger.

According to Ian B. Crosby, partner and lead counsel at Susman Godfrey, which is defending The Times, “these bots compete with the content they are trained on.”

To train generative AI chatbots, artificial intelligence businesses collect publicly available online content, including news stories. Along with a vast array of other human-written items, such as instruction manuals and digital books, the massive language models are also trained on them. As a result, they are better able to answer questions accurately and have a solid grasp of language and syntax.

However, the technology is still in its early stages of development and has numerous flaws. For instance, the Times claimed in its complaint that OpenAI’s GPT-4 damaged the paper’s reputation by erroneously attributing product recommendations to Wirecutter, its product reviews website.

Since the public and business interest in AI technology has grown, especially this year, OpenAI and other AI companies—including rival Anthropic—have gathered billions in financing extremely quickly. Microsoft can benefit from OpenAI’s AI technology thanks to a collaboration with the business.

According to the lawsuit, the Redmond, Washington-based tech behemoth is also OpenAI’s largest supporter, having contributed at least $13 billion to the business since the two started working together in 2019. Under the terms of the deal, Microsoft uses its supercomputers to support OpenAI’s AI research and incorporates the startup’s technology into its goods.

The paper’s complaint coincides with an increase in the number of lawsuits alleging copyright infringement brought against OpenAI. Several authors, including comedian Sarah Silverman, have filed a lawsuit against the business alleging that their works were used without their consent to train OpenAI’s AI models.

A letter accusing the CEOs of OpenAI, Google, Microsoft, Meta, and other AI developers of engaging in exploitative tactics by creating chatbots that “mimic and regurgitate” writers’ words, ideas, and styles was signed by over 4,000 writers in June.

Growing concerns over the application of AI technology have also sparked litigation and labor strikes in other sectors of the economy, including Hollywood. Various stakeholders are aware that the technology has the potential to completely upend their economic model; the challenge is how to react to this realization, according to Sarah Kreps, head of Cornell University’s Tech Policy Institute.

Kreps acknowledged that these chatbots pose a threat to The New York Times, but she also stated that it will be difficult to fully resolve the problem.

She remarked, “There are a ton of other language models out there that are acting in the same way.”

According to the lawsuit, which was filed on Wednesday, Microsoft and OpenAI’s generative AI technologies are closely summarizing information from the Times, even literally reciting it and even imitating its style. Examples of OpenAI’s GPT-4 spitting out huge chunks of Times news pieces were referenced in the complaint.

One such article was an 18-month investigation into New York City’s taxi sector that won the Pulitzer Prize in 2019. Additionally, it referenced outputs from Bing Chat, which is now known as Copilot, claiming to have contained exact quotes from Times articles.

The Times stated that the lawsuit “seeks to hold them responsible for the billions of dollars in statutory and actual damages that they owe for the unlawful copying and use of The Times’s uniquely valuable works,” but it did not specify the precise damages that it is requesting. Additionally, it is requesting that the court force tech companies to erase any AI models or data sets that incorporate its work.

In addition to driving subscriptions to the paper’s online site, web traffic plays a significant role in its advertising earnings. According to the Times, the results of AI chatbots reroute traffic from the newspaper and other copyright holders, decreasing the likelihood that readers will visit the source of the content.

For instance, fewer readers of the Times’ Wirecutter articles click on affiliate links, which reduces revenue for the paper’s product review website.

According to The New York Times, no one has ever been granted permission to utilize its content for generative AI. The action could serve as a means of resuming negotiations to resolve a business issue since it also comes after what appears to be breakdowns in discussions between the newspaper and the two corporations that started in April.

The News/Media Alliance, a trade association that represents over 2,200 news outlets, praised the Times’s action on Wednesday.

“If approached collaboratively, quality journalism and GenAI can complement each other,” stated alliance president and CEO Danielle Coffey. It is not fair use to utilize journalism without permission or payment; it is against the law.

The Associated Press and OpenAI announced a partnership in July that would allow the AI startup to license the news archive of the press organization. This month, OpenAI also inked a comparable agreement with Berlin-based Axel Springer, the parent company of Politico and Business Insider.

As per the agreement, Axel Springer’s media brands will provide summaries of “selected global news content” to users of OpenAI’s ChatGPT. According to the firms, attribution and links to the original publications will be included in the responses to inquiries.

The Times has drawn comparisons between its actions and a copyright case that record companies filed against Napster over 20 years ago, alleging that the file-sharing service had used their music unlawfully. Napster was quickly taken down once the record labels prevailed, yet it had a significant influence on the sector. The music industry is now dominated by streaming.


Leave a Reply