Are article paywalls dead with LLMs? How has nobody built this

For you would have noticed that a lot of news media houses have paywalled articles, especially in The New York Times, WSJ, or The Information. Actively bar websites like 12 feet from accessing their data. 12 feet is still a hit or miss. However, if I ask ChatGPT to explain in detail or summarise an article from a certain website, it tends to do that almost always without any errors. Isn't this a loophole? Is there an AI that can just take in articles from all of these sources, run them through an AI, and then post them out for everyone to read?