"OpenAI 'Stole 10 Million News Articles' to Develop ChatGPT"

Court documents have been released containing claims that ChatGPT developer OpenAI extensively infringed upon the copyrights of news publishers by using over 10 million news articles without authorization to develop its artificial intelligence (AI).

According to AFP on the 17th (local time), the court documents unveiled that day include statements from an insider calling OpenAI's massive use of articles from various news outlets, including The New York Times, for ChatGPT's development an act of theft.

In the released documents, Brent Hecht, the head of applied science at Microsoft—which invested in OpenAI in 2019—pointed out that OpenAI's unauthorized use of over 10 million news articles was "staggering theft on an unprecedented scale" and "the largest labor exploitation in human history."

He also stated that OpenAI may have engaged in "accidental cover-up" activities while reviewing the collected news articles.

According to the court documents, OpenAI used more than 10 million news articles without permission, and about one-third of them were articles from The New York Times.

The New York Times filed a copyright infringement lawsuit against OpenAI and Microsoft in a New York federal district court three years ago.

Joining the lawsuit are U.S. media company Ziff Davis, which owns global IT specialized media outlet CNET, the parent company of the U.S. magazine Mother Jones, investigative journalism news site The Intercept, and various local news outlets across the United States.

OpenAI signed content licensing agreements with various news organizations worldwide when it launched the ChatGPT service in late 2022.

However, concerns have continuously persisted that generative AI software would reduce traffic to internet news sites.

AI companies have attempted to dispel traffic reduction concerns by adding quotes with links to user queries, but even inside the AI industry, claims have emerged that such measures are insufficient.

Reviewing the court documents, OpenAI engineers stated, "No matter how prominently we display links, users will not click them," showing that they were aware the proposed alternatives would be ineffective token gestures.

In response to these claims, OpenAI and Microsoft countered that the use of news content constitutes transformative use that adds value, purpose, and function to the original works, falling under the category of fair use recognized by copyright law.

In early September, the U.S. Department of Justice submitted a brief to the court siding with OpenAI and Microsoft, citing scientific progress, economic growth, and national security.

OpenAI and Microsoft have requested a summary judgment (a ruling for immediate victory) from the court, and if the presiding judge approves it, the lawsuit will not proceed to a full trial.
※ Please note: This article was translated by AI and may contain errors.