The New York Times accuses the company of violating OpenAI copyright law, but the artificial intelligence company notes that the Times is not telling the whole story, but has tricked the model into doing so. “All indications are that they chose their examples from many experiments,” OpenAI said.
OpenAI claims that the Times refused to share examples of its reproduction before the lawsuit was filed. The verbatim examples from the Times appear to be from year-old articles that have been shared on several third-party websites, according to OpenAI. The company acknowledged that it had removed a ChatGPT feature called Browse that inadvertently duplicated content.
However, OpenAI maintained its long-standing position that in order to learn AI models and solve new problems, they need access to a vast pool of human knowledge. He reiterated that while he respects the legal right to own copyrighted works and has offered opt-outs for data inclusion training, he believes that training AI models with data from the Internet falls under fair use rules that allow the reuse of copyrighted works. The company said website owners can block their web crawlers from accessing their data.
The company recently made a similar argument to the UK House of Commons, arguing that no AI system like ChatGPT could be built without access to copyrighted content. It says AI tools must include copyrighted works to “represent the full diversity and breadth of human intelligence and experience.”
But OpenAI said it still hopes it can continue negotiations with the Times for a similar partnership as it did with Axel Springer and the Associated Press. “We look forward to a constructive partnership with The New York Times and respect its long history,” the company said.
Source: The Verge
