In a recent filing, the NYT admitted it does not know how many of its articles were used to train OpenAI's AI models, including GPT-3, as this information was not publicly disclosed. The media outlet is seeking to determine the full extent of ChatGPT's alleged illegal access and infringement. The NYT also argued that a feature in ChatGPT, "Browse by Bing," enabled users to bypass its paywall and access content beyond what was included in its dataset, thereby infringing on the Times' content.
Key takeaways:
- The New York Times (NYT) has denied allegations by OpenAI that it hacked the company's AI chatbot, ChatGPT, to build a copyright infringement case against them.
- NYT is using the lawsuit to investigate the extent of ChatGPT's infringement of their articles and its potential bypassing of the NYT paywall.
- OpenAI maintains that the issues raised by NYT are 'bugs' that will be addressed in future versions of the chatbot.
- NYT has previously banned AI companies, including OpenAI, from using its media archives and data for training large language models, leading to a copyright infringement case against OpenAI and Microsoft.