Meta Faces Backlash After Attempting to Remove Leaked AI Model While Arguing Against Copyright Protections
-
Meta tried to use copyright law to get a leaked version of its Llama AI model removed from GitHub, while arguing that copyright law shouldn't apply when it uses others' content to build AI models.
-
Meta argues that scraping copyrighted text and data to train AI models falls under "fair use" and does not infringe copyrights.
-
Meta's DMCA takedown request to remove the Llama model from GitHub ultimately failed.
-
Microsoft, OpenAI, and Apple argue AI model outputs should be copyrightable, even as they say copyright law shouldn't restrict using others' content as AI training data.
-
Apple argues human-modified AI-generated code should be copyrightable, while Meta thinks its AI models don't infringe copyrights despite reproducing protected works.