A few major newspapers are currently in discussions with OpenAI, the creator of ChatGPT, regarding access to digital news stories. Tech companies have historically used news stories to train their AI models, but newspaper publishers now want a share of the growing market for generative AI. Around 535 news organizations, including the New York Times and The Washington Post, have already installed blockers to prevent their content from being used to train ChatGPT. The current discussions are centered around paying publishers so that the chatbot can provide links to individual news stories in its responses. This would benefit the publishers by providing direct payment and potentially increasing traffic to their websites. Reddit is also reportedly seeking compensation for its data from generative AI companies, and if a deal cannot be reached, it may block search crawlers from Google and Bing. OpenAI recently licensed content from the Associated Press as training data but these talks are more focused on showing news stories in ChatGPT responses. The demand for fair payment for data and copyright lawsuits against AI firms are growing concerns in the industry. OpenAI is negotiating with the newspapers, but discussions are not centered around prior training data, which they argue was obtained legally.
