The New York Times is examining a copyright lawsuit against OpenAI, which could potentially lead to the deletion of the AI language models and a huge fine. This is reported by the US broadcaster NPR, citing anonymous sources familiar with the potential lawsuit. With the passing on of the information, however, pressure could first be built up in the allegedly difficult negotiations between OpenAI and the New York Times regarding the licensing of their content. However, should the lawsuit actually come about, it is likely to become the most sensational dispute over copyright and generative AI models.
Copyright as the sword of Damocles
According to NPR, the New York Times has been negotiating with OpenAI for weeks about financial compensation for allowing the company’s AI models to be trained with texts from the largest US newspaper. However, the negotiations have become so heated that the legal process is now being examined. The main concern at the New York Times is that the OpenAI chatbot ChatGPT and applications based on it could in some way become a direct competitor for the newspaper. Because he creates texts based on the articles in the New York Times and people who read them no longer need to visit the newspaper’s website.
At the same time, the language models of OpenAI – but also other AI companies – have already been trained with texts from the New York Times and other media, for which no permission has been obtained. Courts have not yet answered whether this was even legal, and according to the report, US copyright law offers the newspaper a sharp sword. Should a US federal court rule that OpenAI’s access to the texts of The New York Times was unlawful, it could be ordered that the AI models underlying ChatGPT must be deleted. Then new successor systems would have to be built purely on the basis of approved sources.
In addition, according to NPR, fines of up to $150,000 are possible for “intentional” violations of rights: Since such AI models are trained with millions of texts, this could be “potentially deadly” for companies, says a copyright expert from Vanderbilt University. According to Daniel Gervais, copyright rules are the sword of Damocles that will hang over the heads of AI companies for years to come. So they had to negotiate a solution. So the pressure to reach an agreement in the negotiations is probably greater for OpenAI than for the New York Times. With the threat of a lawsuit, the newspaper is probably tightening the thumbscrews. Just a few days ago, she banned training with her content.
#Copyright #York #Times #Considers #Potentially #Fatal #Lawsuit #OpenAI