The U.S. Copyright Office just released a “pre-publication” version of its series on AI and Copyright. The Register of Copyrights pre-publication report, entitled, Generative AI Training, contained the statement that it was “released in response to congressional inquiries and expressions of interest from stakeholders. A final version will be published in the near future, without any substantive changes expected in the analysis or conclusions.”
The draft report (summarized here) has some controversial opinions on AI training and copyright infringement including about how Generative AI systems are trained, what acts involved in training may implicate the reproduction right, what parts of a work contain protectable expression, whether training AI systems is transformative, how the fair use factors should be applied including how factor four (effect of the market) should be applied, and how to weigh the public interest with fair use factors.
Whether the report will ever be released is unknown. Right after the pre-publication version of the report was released, President Trump fired the Register of Copyrights, Shira Perlmutter. This followed President Trump’s firing of the Librarian of Congress, Carla Hayden on Thursday. At this point we don’t know whether the termination of Shira Perlmutter was in response to the report, or part of a larger movement by conservative activists to eliminate supposed partisans from the U.S. government, or some other reason.
We also do not know what prompted the release of a pre-publication version of the report. Was it to influence the two summary judgment motions currently before thee U.S. courts in the ThomsonReuters and Meta cases, or to ensure the report was released with knowledge that it could be buried if the Register of Copyright’s employment was terminated, or some other reason?
The report Generative AI Training follows its prior report on copyright eligibility for generative AI content, summarized here.
1 comment