<p>Executives at OpenAI and Microsoft acknowledged that artificial intelligence poses an "existential threat" to news publications as the New York Times sues the companies over copyright infringement.</p>
<p>Brent Hecht, Microsoft’s director of applied science, expressed concerns about the negative effects that AI models could have on the internet and the company’s business. In an internal document, Hecht stated, "Our AI content strategy has started a ‘doom loop’ that will hurt the performance of our models and the entire web at the same time." He noted the unusual situation where an end product threatens the economic foundations of its essential suppliers.</p>
<p>The New York Times contends that chatbots, including OpenAI’s ChatGPT, should not utilize its copyrighted content for training, as the use of news articles diverts users from engaging with their publications.</p>
<p>Microsoft CEO Satya Nadella remarked that chatbots have "substituted" traditional journalism by providing information directly on AI platforms instead of directing users to the original sources.</p>
<p>Nick Turley, vice president and head of ChatGPT, stated that publishers face an "existential threat" from AI chatbots, which he described as "largely substitutive" and likely to become even more so as they improve.</p>
<p>Hecht referred to the trend of chatbots accessing news articles for training as "an astonishing theft of unprecedented proportions," suggesting it could represent the "largest theft of labor in human history." These statements may strengthen the plaintiff’s case against allowing chatbots to train on copyrighted works without permission or compensation.</p>
<p>The New York Times operates a paywall for its articles, but ChatGPT can bypass subscription requirements. OpenAI President Greg Brockman previously mentioned efforts to circumvent the news outlet’s paywall to "scrape" its content, according to court documents.</p>
<p>The documents also identify custom GPTs capable of extracting information from paywalled articles, such as "Bypass Paywall" and "Article Reader." OpenAI lacks a licensing agreement with the New York Times, leading to accusations that it harms the newspaper’s business model by diverting readers. However, OpenAI has licensing agreements with other news publishers.</p>
<p>This information regarding AI leaders’ views on these legal issues follows a statement of interest filed by the Department of Justice, which supports the defense presented by OpenAI and Microsoft. The DOJ argued that restricting AI models from training on copyrighted material due to a "misunderstanding" of fair use could hinder creative and scientific progress, as well as American economic mobility.</p>