4 ms·

(not making a argument for either side) I believe the Google Books decision is probably the strongest precedent for this in the US: https://en.wikipedia.org/wik

• maronato 3y ago
  The conclusion of the appeal reads: > Google's unauthorized digitizing of copyright-protected works, creation of a search functionality, and display of snippets from those works are non-infringing fair uses. The purpose of the copying is highly transformative, the public display of text is limited, and the revelations do not provide a significant market substitute for the protected aspects of the originals. Google's commercial nature and profit motivation do not justify denial of fair use. IMO this is very different from what OpenAI is doing. - The purpose of Google books was to search for content and then use the results of the search to buy the books from the original publishers. OpenAI doesn’t attribute results nor provides a way to pay for the content it returns. - Google books doesn’t provide the entire book as a response from the query, only the snippet with the result. OpenAI does return entire articles verbatim or paraphrased as result of queries. - Users of Google Books do not use it in place of actual books. ChatGPT replaces the need to read the actual articles/sources.