5 hours ago

Trump Fired Copyright Chief After She Warned Of AI Companies Violating The Law

The Trump administration quietly fired the head of the U.S. Copyright Office over the weekend after her office issued a lengthy report condemning artificial intelligence companies for training their models on copyrighted works, often without the copyright owners’ consent or compensation.

“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,” the third part of the 113-page pre-publication report concluded.

Register of Copyrights Shira Perlmutter had held the position since President Donald Trump’s first term in October 2020, operating under the Librarian of Congress, Carla Hayden.

Hayden herself was abruptly fired last week without cause in a two-sentence email. Trump on Monday tapped Todd Blanche, his former personal attorney, to run the Library of Congress despite having no apparent prior professional experience in the field.

Both Perlmutter’s and Hayden’s firings would appear illegal, since both operate under the purview of Congress ― not the president.

Trump himself didn’t appear to understand Perlmutter’s firing. He confirmed the news by sharing a social media post written by Mike Davis, a Republican attorney, that condemned it as “100% unacceptable.”

“Now tech bros are going to attempt to steal creators’ copyrights for AI profits,” griped Davis.

Rep. Joe Morelle (D-N.Y.), the top Democrat on the Committee on House Administration, highlighted Elon Musk’s connection to the firing.

“It is surely no coincidence [Trump] acted less than a day after [Perlmutter] refused to rubber-stamp Elon Musk’s efforts to mine troves of copyrighted works to train AI models,” Morelle said in a statement.

Democratic U.S. Senators Adam Schiff of California and Chuck Schumer of New York also condemned the firing as unlawful in a joint statement.

Companies such as OpenAI, Meta, and Elon Musk’s xAI (which acquired X, formerly Twitter, earlier this year) are predicated on vast troves of data, not all of it acquired legally.

Court records show Mark Zuckerberg’s Meta, for instance, illegally downloaded nearly 82 terabytes of pirated books to train its AI model, LlaMA.

Last year, an OpenAI whistleblower who alleged the company was built on copyright violations was found dead of an apparent suicide.

A 2023 class-action suit by Sarah Silverman and other authors against Meta and ChatGPT parent OpenAI accuses the companies of being “industrial-strength plagiarists.”