OpenAI finds lawsuit filed by The New York Times ‘unfounded’

OpenAI has responded to the complaint against the company by The New York Times. The AI ​​company states that the American newspaper ‘does not tell the full story’. OpenAI therefore believes the lawsuit that the newspaper filed at the end of December 2023 is ‘unfounded’.

The ‘burps’ caused by The New York Times are according to OpenAI from ‘years old articles’ that have been distributed on various websites. OpenAI says it appears that prompts were “intentionally manipulated.” When reproducing the prompts, the models would not respond as The New York Times insinuated, the ChatGPT creator claims. The blog post states that OpenAI suspects that The New York Times has selectively chosen examples that support the newspaper’s claim, also known as cherry picking.

In late December, The New York Times sued OpenAI and Microsoft for violating the newspaper’s copyright. The newspaper claims that the tech companies have misused ‘millions’ of articles from The Times to train their chatbots. OpenAI previously said it was in discussions with the newspaper. Although the talks did not lead to solutions, they are said to be of a ‘constructive’ nature. The lawsuit was therefore a surprise for OpenAI.

Furthermore, OpenAI states that training AI models using publicly available material on the internet is considered ‘fair use’, which would be based on ‘widely accepted precedents’. Nevertheless, OpenAI offers an opt-out for publishers, which prevents the tech giant’s tools from accessing the publisher’s sites. The New York Times reportedly opted out in August 2023.