The New York Times is suing OpenAI and Microsoft for using copyrighted material

loading

I The New York Times File a lawsuit against to OpenAI and Microsoft, accusing the two companies of using his articles, which are protected by copyright, to develop their own artificial intelligence systems. So-called generative AI, such as ChatGPT developed by OpenAI, uses large amounts of written text and images to develop systems capable of producing new text and images. according to The New York Timeswhich is one of the most widely read newspapers in the world, is said to have artificial intelligence systems trained on its texts stealing visitors from its site.

The lawsuit does not specify the amount of compensation the newspaper is seeking, but it says the companies will be liable for “billions of dollars” in damages, and calls for the destruction of software based on models trained on copyrighted material. According to a complaint The New York TimesIn some cases, ChatGPT may produce almost identical word-for-word transcripts of articles published on its site and can only be accessed if you subscribe to the newspaper for a fee.

This isn't the first lawsuit filed against OpenAI: in September several novel authors, including George R.R. Martin, John Grisham, and Jonathan Franzen, filed suit as well, for similar reasons. The New York TimesBy November, many essayists had done the same. According to what he claimed The New York Times In itself, the complaint follows negotiations that began in April between the newspaper and Microsoft (which has very close ties to OpenAI), which yielded no results.

– Read also: For AI to work, human intelligence needs to do very boring jobs

On the contrary, the German publisher Axel Springerwhich publishes among others BildThe most widely circulated newspaper in Germany, e POLITICOAnd the American News Agency News agency It has reached agreements with OpenAI. The AP granted access to part of its text archive to ChatGPT. The agreement with Axel Springer instead stipulates that ChatGPT can provide users with excerpts of news published in the group's newspapers. In both cases, the economic scope of the contracts was not revealed.

One of the biggest concerns about generative AI programs is so-called “hallucinations,” situations in which the program essentially inserts false information into its response to the user, sometimes attributing it to a source like a newspaper. In fact, the program works thanks to a statistical model that tries to guess which words make sense to use in a given context: this sometimes generates correct answers according to linguistic rules, but without any commitment to reality. the The New York Times He claims this could represent damage to his reputation.