President Donald Trump has removed Shira Perlmutter from her role leading the U.S. Copyright Office, placing a key copyright official at the center of the intensifying fight over artificial intelligence, fair use and the value of creative work.
The firing, reported by CBS News and Politico, drew an immediate political response because it followed the release of a U.S. Copyright Office report on copyright and artificial intelligence. That report did not decide individual disputes, but it did draw boundaries around how AI companies may be able to use copyrighted material when training models.
Why The Firing Matters
Perlmutter took over the Copyright Office in 2020, during the first Trump administration. She was appointed by Librarian of Congress Carla Hayden, whom Trump also fired this week.
The Copyright Office sits in a crucial position as AI companies, creators and courts wrestle with whether copyrighted works can be used to train AI systems without permission. Its reports do not resolve every lawsuit, but they can shape how policymakers, companies and rights holders understand the legal landscape.
Representative Joe Morelle, the top Democrat on the Committee on House Administration, described the removal as politically and legally alarming. In his statement, he said, "Donald Trump’s termination of Register of Copyrights, Shira Perlmutter, is a brazen, unprecedented power grab with no legal basis."
Morelle also tied the timing directly to AI training. He said, "It is surely no coincidence he acted less than a day after she refused to rubber-stamp Elon Musk’s efforts to mine troves of copyrighted works to train AI models."
The AI Report At The Center Of The Dispute
Morelle pointed to a pre-publication version of a U.S. Copyright Office report released this week. The report focuses on copyright and artificial intelligence and is part three of a longer report.
The report’s central point is cautious but significant: the Copyright Office says it is "not possible to prejudge" how individual cases will come out. At the same time, it says there are limits to how much AI companies can rely on "fair use" when they train models on copyrighted content.
That distinction matters. The report indicates that some uses, such as research and analysis, would probably be allowed. But it takes a much harder view of commercial AI training that uses large bodies of copyrighted work to create expressive outputs that compete in existing markets.
The report says that "making commercial use of vast troves of copyrighted works to produce expressive content that competes with them in existing markets, especially where this is accomplished through illegal access, goes beyond established fair use boundaries."
For creators and copyright holders, that language supports the argument that AI training cannot be treated as automatically protected simply because it involves a new technology. For AI companies, it signals that fair use may remain available in some contexts, but not as a blanket answer to every training practice.
What The Copyright Office Wants To See Next
The report does not call for immediate government action. Instead, the Copyright Office says intervention "would be premature at this time." That position leaves room for courts, licensing talks and market behavior to keep developing before lawmakers or agencies step in more forcefully.
At the same time, the report points toward licensing as a likely part of the future. It expresses hope that "licensing markets" where AI companies pay copyright holders for access to content "should continue to develop."
The report also says "alternative approaches such as extended collective licensing should be considered to address any market failure." In plain terms, the office is suggesting that if direct licensing does not work well enough, broader systems for managing rights and payments may need attention.
The report therefore lands between two extremes. It does not say every AI training use is unlawful, and it does not say AI companies should be free to use copyrighted work without constraint. Instead, it frames the issue as one where purpose, market impact, access and competition all matter.
The Bigger Fight Over AI And Copyright
The firing comes as AI companies, including OpenAI, face a number of lawsuits accusing them of copyright infringement. OpenAI has also called for the U.S. government to codify a copyright strategy that gives AI companies leeway through fair use.
Musk is part of the broader context because he is a Trump ally and a co-founder of both OpenAI and xAI. The source article also notes that xAI is merging with X, formerly Twitter.
Trump alluded to the news on Truth Social by "ReTruthed" a post from attorney Mike Davis linking to the CBS News article. Davis’ own message added a twist, because he appeared to criticize the firing while warning that "tech bros are going to attempt to steal creators’ copyrights for AI profits."
Musk has also recently expressed support for Square founder Jack Dorsey’s call to "delete all IP law." That stance sits uneasily beside the Copyright Office report’s emphasis on fair use limits, licensing markets and the possibility of other licensing approaches if the market does not function properly.
The immediate story is a personnel decision. The larger issue is whether the rules for copyrighted works can adapt to AI training without either cutting creators out of the market or blocking research and analysis that the Copyright Office says would probably be allowed. Perlmutter’s removal makes that debate more political, but the underlying legal and commercial questions remain unresolved.