Advertisement
Microsoft has launched a defense against The New York Times’ lawsuit targeting OpenAI, alleging that the claims made by the publisher are “unsubstantiated” and likening the legal battle to past conflicts over technological advancements.
In a motion filed on Monday seeking to dismiss part of the lawsuit, Microsoft argued that The New York Times presented a skewed narrative of “doomsday futurology,” drawing parallels to Hollywood’s resistance to the VCR in the 1970s.
The lawsuit centers on concerns that OpenAI’s ChatGPT chatbot could disrupt the news industry by generating content that could potentially decimate traditional journalism. Microsoft, as OpenAI’s largest investor with an investment of approximately $13 billion, has a vested interest in the outcome of the case.
The legal dispute highlights the growing tensions between OpenAI and the media sector, particularly regarding the use of copyrighted content to train language models. The Times accuses OpenAI and Microsoft of copyright infringement and misappropriation of its intellectual property in training LLMs.
OpenAI, in response, has alleged that The New York Times paid individuals to manipulate ChatGPT to produce examples of copyright infringement. Microsoft’s lawyers argue that the use of copyrighted content to train LLMs does not directly compete with the market for the original works but rather facilitates language modeling.
Ian Crosby, lead counsel for The New York Times, countered Microsoft’s arguments, stating that the lawsuit exposes the unauthorized use of copyrighted works by Microsoft and OpenAI. He emphasized that The Times discovered evidence of copyright infringement and holds the defendants accountable for their actions.
Despite the legal battle, OpenAI has surged in prominence since releasing ChatGPT to the public, achieving a valuation exceeding $80 billion. CEO Sam Altman has defended OpenAI’s practices, stating that the company does not necessarily require access to specific datasets to develop its AI models effectively.
While OpenAI has acknowledged the necessity of copyrighted materials in training AI models, it has also pursued partnerships with media companies like Axel Springer, CNN, Fox Corp., and Time to license their content, signaling a potential shift in its approach to data acquisition.