What might mean the first ruling of artificial intelligence in the United States of IP law

Publishing rights claims against artificial intelligence companies have obtained a possible batch.

An American federal judge last week She issued a brief judgment In the case of Thomson Thomson Reuters Thomson Tostented against Ross Intelligence. The judge found that Ross’s use of Reuters content to train the legal research platform for Amnesty International violates the intellectual property of Reuters.

The result can have effects on more than 39 Amnesty International Claims related to copyright They are currently working on their way through American courts. However, it is not necessarily peace without you for the prosecutors who claim that artificial intelligence companies violate their IP rights.

Everything about headnotes

Ross was accused of using Headnotes – Legal Decisions – from Westlaw, Reuters legal research service, to train artificial intelligence. Ross has marketed artificial intelligence as a tool for analyzing documents and performing searches based on inquiries through court files.

Ross had argued that his use of copyright -protected headscarf was legally defended because it was transformed, which means that he had replaced the head to serve a different job or market significantly. In his brief rule, Stefanos Pipas, the judge who heads the case, did not find this special argument.

In his opinion, Ross said that Ross was re -filling out the Westlu domains in a way that directly repeated the legal research service in Westlu. The starting starting platform did not add meaning, purpose, or new comment, as the BIBAS was identified – the Ross claim was undermined by the transformative use.

In his decision, Pepas also referred to Ross’s commercial motives as a reason for the start -up defense of the brand. Ross sought to take advantage of a competing product directly with Westlaw, and without a large “allocation” of the IP protected Westlaw materials.

Chop Josh, a professor of the University of Sri Lips, who studies the intellectual property law, launched a “strong victory” by Reuters Thompson.

“The trial will continue (but) Thompson Reuters obtained a brief ruling, a victory at this stage of litigation.” “The judge also emphasized that Ross is not entitled to obtain a brief ruling on his defenses, such as fair use and integration. As a result, the case continues to trial with a strong victory on Reuters Thompson.”

Tight in the application

Indeed, there is at least one group of prosecutors in the case of other publishing rights of artificial intelligence The court asked to consider Pepas’s decision. But it is not yet clear whether a precedent will affect other judges.

BIBAS’s opinion has made clear the point of distinction between “AI Tolidi” and AI that Ross used, which did not generate content but only spit the judicial views that were already written.

Trucific artificial intelligence, which is located at the Center for Copyright Cases against Companies such as companies Openai and MidjourneyIt is frequently trained on huge amounts of content from public sources around the web. When many examples feed, the AI ​​Agency can generate speech, text, photos, videos, music and more.

Most companies that develop artificial intelligence argue that Doctrines of just use Freeze their practice to ignore data and use training without compensation – or even credit – data owners. They argue that they are entitled to use any content available for the public for training and that their models are valid to produce transformational work.

But not every copyright holder agrees. It refers to this phenomenon known as the name SackingAs artificial intelligence creates a closely trained work.

Randy McCarthy, US Patent Lawyer in the Law Hall, said that Pipas’s focus on “the effects of the market for original work” may be a key to rights -holder’s cases against artificial intelligence developers. But he also warned that Pepas’s opinion is relatively narrow and that it might turn at the appeal.

“There is one clear thing, at least in this case: just the use of copyrights protected as training data (for) AI cannot be said that it is considered fair in itself.” “(But) is one battle in a greater war, and we will need to see more developments before we can extract from this law related to the use of copyrights as copyright as artificial intelligence training data.”

Another techcrunch lawyer spoke with Mark Lezama, a vowing partner in Knobbe Martens, who focuses on patent disputes, believes that Bibas’s opinion can have wider effects. It is from the point of view that the judge’s thinking can extend to artificial intelligence in its various forms.

He said: “The court refused to defend fair use as a partial legal issue because Ross used Headnots to develop a competitive legal research system.” “Although the court has hinted that this may be different from the situation that includes the artificial intelligence, it is easy to see a news site claiming that copying its articles to train artificial intelligence is not different because the artificial intelligence of the user’s interest.”

In other words, publishers and copyrights who write them with artificial intelligence companies have a simple reason for optimism after the decision – focus on slight.