The Chamber of Progress, a self-styled “progressive” industry trade group supported by most of the biggest tech platforms, has urged the Trump administration to intervene in a litany of copyright cases involving artificial intelligence firms, to try to stop authors and publishers from having their work used for training AI models without permission.
The pleading comes as Anthropic prepares to pay authors $1.5 billion, the largest award in the history of copyright law, for pirating their work, in a settlement announced last month. OpenAI, Microsoft, Google, and Meta are named defendants in the more than 50 active lawsuits over AI intellectual-property theft.
In a letter to Michael Kratsios, the lead science adviser to President Trump, the Chamber of Progress estimates that AI companies could be liable under the Copyright Act for up to $1.5 trillion for stealing copyrighted work on which to train their models. The letter’s authors claim that this represents “an existential risk” to AI companies, and that the cases should be tossed out under a “fair use” standard.
The Chamber of Progress’s campaign to promote fair use, which they have created a campaign around called “Generate and Create,” comes as at least three of the nonprofit organization’s past or current backers are being sued over copyright claims: Meta, Google, and the AI art generator Midjourney. Another current funder, Nvidia, relies heavily on AI development for its continued success, and venture capital firm a16z, with several AI startups in its portfolio, also funds the nonprofit.
Even if training of AI models might be fair use, mass downloading of pirated content is not.
Google in particular would be quite interested in such a request for intervention. The Chamber’s CEO, Adam Kovacevich, is a former Google lobbyist; several Chamber colleagues and board members worked with him at Google; and leaked emails show ongoing coordination between the tech giant and the lobbying group. The lawyer defending Google in a 2024 lawsuit over its AI image generator co-authored a Chamber amicus brief in litigation involving copyrighted music lyrics.
The Chamber of Progress did not respond to a request for comment.
AI companies seeking to reduce their trillion-dollar legal exposure have argued that their models are merely reading books and articles, and that this fair use should not trigger copyright violations. Of course, AI companies are not going to bookstores and purchasing the books that are fed into the models. They are scraping the web and pirating mass quantities of data on an automated basis. In unredacted court filings, Meta acknowledged using Library Genesis (LibGen), a “shadow library” of pirated material.
“The cases advance two theories, one of which I agree with, one of which I don’t,” said Cory Doctorow, author of Enshittification and a copyright expert going back to his long tenure with the Electronic Frontier Foundation. “They advance the theory that making temporary copies of works, analyzing them, and publishing the analysis is a copyright infringement. I think that’s wrong. They also advance the theory that pirating e-books infringes copyright. That’s obviously right.”
In the Anthropic case, Judge William Alsup made this distinction. Even if training might be fair use, mass downloading of pirated content is not. That opens AI firms up to significant damages. Each violation of the Copyright Act can cost up to $150,000 per copyrighted work, per the statute, and while violations of the Digital Millennium Copyright Act are smaller, they are based on each “act of circumvention.” The logic applies to online publications, books, music, art, and countless other forms of content.
Vacuuming up pirated content by the petabyte is almost the definition of willful copyright infringement. But the Chamber of Progress claims that the overall legal regime would “radically prejudice AI developer defendants” and act as “a de facto tax on AI innovation.” Doctorow said that the Chamber was “trying to conflate” the two theories of harm in an effort to defend AI firms from piracy charges.
The Chamber adds that the statutes are “outdated” because AI is sucking up more copyrighted works than the law’s authors could have possibly conceived of in the late 20th century. This puzzling argument suggests that wide-scale violations of a statute should lead to new laws to immunize the violators.
To that end, the Chamber of Progress asks President Trump to issue an executive order directing the Justice Department and the Commerce Department to file amicus briefs or intervene in AI copyright cases, to actively oppose efforts to hold AI companies accountable for pirating content. It even suggests that the Trump administration should try to block class action certification (even as all potential plaintiffs are seeing their data pirated and used in exactly the same way) and that damages under copyright law would violate the due process clause of the Constitution.
If intervening in the courts doesn’t work, the Chamber of Progress asks the White House to draft legislation to protect AI companies that Congress can take up in 2026, to limit damages under the Copyright Act and the DMCA. The last time that Congress proposed legislation to deregulate the AI sector, it failed in the Senate 99-1.
THE CHAMBER OF PROGRESS, WHICH SEES ITSELF as a democratic organization, received $29.9 million in revenue in 2022 and $8.6 million in 2023, the last years where it has submitted public tax returns. Its “corporate partners” include a16z, Airbnb, Amazon, Apple, Coinbase, Google, Intuit, Midjourney, Nvidia, Ripple, and Uber; Meta and Twitter were past partners at launch. At least seven of the Chamber’s corporate partners have donated to Trump’s ballroom project.
Though the Chamber shares its lead outside lobbyist, Steve Elmendorf, with Meta and Amazon, Google appears to be a driving force at the organization, given its CEO Kovacevich and several other ties. (Elmendorf, in fact, previously represented Google.) Google also happens to be fighting one AI copyright lawsuit over six of its products (narrowed from what was initially 16). Another suit over the AI summaries in Google search was filed in September by Penske Media, owner of Rolling Stone and Billboard.
The Chamber has been active on AI copyright issues for a few years, going back to rebutting comments filed with the Copyright Office by the Federal Trade Commission in 2023. Last year, it co-led an amicus brief (along with right-leaning pro–Big Tech organ NetChoice) on behalf of Anthropic in the case that the maker of AI chatbot Claude ultimately lost. One of the authors of that amicus, Eric Tuttle of Wilson Sonsini, is currently representing Google in the consolidated AI copyright suit over six of its products.
The Chamber co-signed a coalition letter in late 2023 endorsing minimal regulation for AI regarding copyright, which was written by copyright attorney Sy Damle, one of the counsels for the defendants in the AI copyright case involving Midjourney. Damle also represented Anthropic in its losing case, along with several AI startups in copyright infringement cases.
Jess Miers, who until last year was the Chamber’s in-house attorney, previously worked on Google’s public-policy team. As the organization’s copyright counsel, she worked on the Generate and Create campaign that was launched last year, and led efforts to defend member organization Midjourney, which has been sued by artists on copyright grounds. “We’re fighting back” against authors and artists, the Chamber’s Generate and Create campaign messaging states. Miers left to teach at the University of Akron shortly after the campaign launch, but the Chamber of Progress’s campaign against Big Author evidently continues apace.