Home NETWORK POLITICS “AI really bad for industry”: AI company first loses in American legal...

“AI really bad for industry”: AI company first loses in American legal dispute

0


In the United States, a media group first reinforced against the AI ​​group which used its content to train its algorithm. If other courts are involved in the assessment, it would be “really bad for the generic AI industry,” wired for IT law. It was now decided to have a legal dispute between Thomson Reuters and long-awaited Start-up Ross Intelligence, which was not at all about generative AI, presented in early 2020. In many other legal proceedings, it was the question of what was training. An AI with non -liched material comes under appropriate use. This is the one that was now definitely denied.

Advertisement


Federation Digital Gharana included 60 billion euros in four years

In this case, Westlav, a subsidiary of Thomson Reuters. Among other things, it maintains the database with sleep -headnotes or orientation rates, which are small summations of legal texts, such as court decisions. As a contestant, Ross Intelligence wanted to create an AI-based search engine for these short texts, Richter Stephenos summarized Bibas with delawareFor training, the start-up wanted to license the contents of Westlav, which was rejected by the group. As a result, Ross Intelligence of another company, which eventually created training data for Westlav’s content.

When the AI ​​company was drawn to court in 2020, it defended the process based on fair-use theory. Under some rules, it enables unauthorized use of copyright content and is also presented for defense in other cases against AI companies. In the specific case, however, it was not accepted, with which the responsible judge also canceled a separate initial decision. The main decisive factor was that a competitive product was created. The decision states “None of the potential defense of Ross is valid – I reject them”. “Mimiting our content was not a ‘proper use’,” Thomson Reuters are satisfied,

For many pending legal proceedings against AI companies, now the decision does not mean anything good. If other recipes reject the context of appropriate use equally, a decisive base will be abolished to train models with immense quantities of materials. IT legal expert James Greecimelman from Cornell University References to wired The fact is that Bibas has rejected the most court decisions, in which AI companies have so far refused as “irrelevant”. Ross intelligence was the result of a legal dispute Already closed in 2021 Made.


(MHO)

AI-Janit goes viral after a video: Scarlett Johansson demands regulation

NO COMMENTS

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Exit mobile version