From Daily Mail:

The New York Times filed a copyright lawsuit against OpenAI and Microsoft, alleging the tech companies illegally used their articles to train chatbots. OpenAI retaliated, accusing The Times of manipulating prompts to create misleading evidence and dismissing the hacking claim as false and irrelevant.

The New York Times denied hacking OpenAI’s systems and refuted claims of misleading evidence for the lawsuit. The newspaper accused OpenAI of copying their content on a massive scale and attempting to discredit the suit by using the term ‘hacking.’ OpenAI argued that using publicly available internet materials, like The Times’ articles, is fair use.

The New York Times argued that OpenAI and Microsoft exploited their articles to develop artificial intelligence programs, emphasizing that there was nothing transformative about using their content without permission or payment. The lawsuit cited examples of AI ‘hallucinations’ or false information attributions, which OpenAI claimed were intentionally manipulated by The Times.

OpenAI responded by stating that The New York Times induced regurgitations from their models through manipulated prompts with lengthy excerpted articles, contrary to the usual behavior of their models. OpenAI emphasized that intentionally manipulating their models to regurgitate was not appropriate and against their terms of use.

The New York Times highlighted the financial gains of the tech companies from using their work, despite attempts to negotiate fair compensation. Microsoft’s market capitalization increased by a trillion dollars, and OpenAI’s valuation rose to $90 billion due to their deployment of Times-trained language models. OpenAI expressed hope for a constructive partnership with The Times and continued collaboration with news organizations.



Read more at Daily Mail: New York Times denies OpenAI’s claims it HACKED their systems by using ‘hired gun’ to create misleading evidence