Agreed, they could carefully coerce the model to more or less output some of their articles, but the premise that users were routinely doing this to bypass the paywall is silly.

▲

terminalshort 8 hours ago | parent [-]

Especially when you can just copy paste the url into Internet Archive and read it. And yet they aren't suing Internet Archive.

	▲	realusername a minute ago \| parent \| next [-]
		Let's be real, they are suing OpenAI because they have way more money than the Internet Archive and they would be happy with a cut
	▲	acdha 7 hours ago \| parent \| prev [-]
		Copyright law isn’t binary and has long-running allowances for fair use which take into consideration factors like scale, revenue, and whether it replaces the original. As a real non-profit, the Internet Archive is not selling its copies of the NYT and it’s always giving full credit to the source. In contrast, ChatGPT does charge for their output and while it may give citations that’s not a given.