>>solard+Aj
I think the exact opposite is true: as long as AI depends critically on scrupulous news media to be able to generate info about current events, it is far more important to protect the news media than the AI training models. OpenAI could survive even if it had to pay the NYT for redistributing their works. But OpenAI can't survive if no one is actually reporting news fairly accurately. And if the NYT were to go bankrupt, all smaller players would have gone under looooong before.
In some far flung future where an AI can send agents to record and interpret events, and process news feeds and others to extract and corroborate information, this would greatly change. But probably in that world the OpenAI of those times wouldn't really bother training on NYT data at all.