The OpenAI-Micro lawsuit in a digital news publication alleges reproduced journalism by chatbots trained on article excerpts without compensation or permission

The lawsuit echoes many of the same claims made by other news organizations in suing OpenAI and Microsoft. The New York Times alleges that the newspaper’s journalism was reproduced in some way. Digital news sites The Intercept, AlterNet, and Raw Story filed their own legal action in February.

The publications, all owned by the hedge fund Alden Global Capital, claim that both OpenAI and Microsoft trained on their articles without compensation or permission. The plaintiffs included as evidence several excerpts from conversations with both ChatGPT and Copilot showing that both chatbots reproduced lengthy excerpts of specific articles on command, indicating that their training datasets included the texts of those articles.

They also showed screenshots of Copilot, which can search the web in real time, reproducing entire news articles verbatim a day or two after those articles were posted, without “a prominent hyperlink” back to the original article. The companies also claim that chatbots often attribute false facts or hallucinations to publications.

Companies that use copyrighted material for their artificial intelligence models have to obtain the publishers’ consent and pay fair value for such use.

In addition, according to the suit, ChatGPT at times falsely attributes reporting to the newspapers in the answers it generates, tarnishing the reputation of the news outlets.

In another example noted by the suit, ChatGPT was asked if smoking cures asthma, and the chatbot fabricated that The Denver Post published research indicating that smoking can be a cure for asthma. That assertion is obviously false, and the paper never published such research.

“This issue is not just a business problem for a handful of newspaper or the newspaper industry at large,” lawyers for the newspapers wrote in the suit. It’s a major issue for civil life in America.

Eight newspapers sue OpenAI, Microsoft for copyright infringement: A challenge to the publishers and the news publishing giant, Google, Apple, Amazon, and Google

“Along with our news partners, we see immense potential for AI tools like ChatGPT to deepen publishers’ relationships with readers and enhance the news experience,” said a spokesperson for OpenAI.

Attorneys for the newspapers, which are all owned by a New York investment fund, are asking for millions of dollars in damages, for using their copyrighted work to end.

It has been said that it would be nearly impossible to rebuild the models without completely incorporating works published by the newspapers.

Gary Marcus is the author of a book entitled “Taming Silicon Valley: How We Can” and he said it would cost a hundred million dollars to re-train a model. Yes. That AI Work For Us.

Marcus said that it can be hard to remove copyrighted materials from a dataset due to the fact that some websites often hold versions of copyrighted stories in their posts.

A legal battle is set to be high-stakes and involves one of the world’s leading artificial intelligence companies against news publishers, which experts say isn’t certain about the law.

Yet OpenAI has long claimed that its so-called “large language models,” hoover up vast amounts of data from all corners of the internet under what is known as the “fair use” doctrine.

If a work is substantially changed, it can be used without the permission of the copyright holder if it does not compete with the original.

Source: Eight newspapers sue OpenAI, Microsoft for copyright infringement

Is OpenAI Responsible for Editorial Content? Why News and Observations are Hating the OpenAI Law, and How it is Trying to Prevent It

It will likely take years of court battles and a long appeals process to determine if leading technology firms like OpenAI have violated the law because legal scholars say it is not certain that the law is on the side of artificial intelligence companies.

Other publishers have chosen to take a more conciliatory path with the company. The Financial Times and The Associated Press have struck agreements with OpenAI to pay for the use of copyrighted material.