Oct 29, 2025 4 min read

After Anthropic’s $1.5 billion settlement, big tech takes aim at statutory damages in copyright disputes, seeks Trump’s support

Last month Donald Trump asked for more submissions on AI regulation. One lobby group has used its submission to not only argue that AI training is ‘fair use’, but to also criticise statutory damages in US copyright law which it says are unfairly forcing AI companies into billion dollar settlements

After Anthropic’s $1.5 billion settlement, big tech takes aim at statutory damages in copyright disputes, seeks Trump’s support

Lobbyists for the tech sector are now turning their attention to the statutory damages that are available under US copyright law, which allow copyright owners to claim up to $150,000 every time a copyright is infringed. That can result in mega-damages if, for example, an AI company is found liable for infringing the copyright in millions of works copied during AI training processes.  

The new lobbying focus is a direct result of the legal battle between Anthropic and a group of book authors, in which the AI company was pressured into a quick $1.5 billion settlement because it potentially faced having to pay more than a trillion dollars in statutory damages. 

Campaign group Chamber Of Progress is now calling on President Donald Trump to intervene, and oppose excessive statutory damages in court and reform this element of copyright law in Congress. 

“For the United States to maintain leadership in AI innovation”, it writes, Trump’s government should “urgently seek to rationalise the present legal environment for generative AI development”, first by “engaging” with all the AI copyright cases currently working their way though the US courts, and then by “designing and advancing innovation-compatible reform of the current statutory damages framework”.

Trump’s government opened another consultation on the regulation of AI last month. It follows the publication of Trump’s big ‘AI Action Plan’ back in July, which didn’t really say anything about the copyright obligations of AI companies, despite there now being dozens of US lawsuits between AI companies and copyright owners, including record labels and music publishers.   

However, in the midst of the customary waffle Trump delivered when launching that plan he said “you can’t be expected to have a successful AI programme when every single article, book or anything else that you’ve read or studied, you’re supposed to pay for”. Having AI companies secure licences for every copied work just “isn't doable" he then declared. 

The copyright industries, including the music industry, would obviously strongly dispute that view. However, Trump’s rambling remarks are heavily cited in new submissions to the US government by both the Computer & Communications Industry Association and the Chamber Of Progress, a campaigning organisation that counts Amazon, Apple, Google, Midjourney and music AI company Suno as ‘partners’.

Most AI companies argue that AI training is ‘fair use’ under US copyright law meaning they don’t need permission from copyright owners when copying existing content onto their servers in order to train generative AI models. Copyright owners disagree and that dispute is at the heart of all the lawsuits. The CCIA and Chamber Of Progress reassert their fair use arguments in their new submissions.

In the Anthropic case the judge said AI training was fair use, but only if the AI companies curated their training data from legitimate sources. 

Anthropic had downloaded millions of ebooks from piracy sites. Which meant it was probably liable for copyright infringement in relation to those pirated books. Apply statutory damages of $150,000 per infringement and Anthropic was facing a potential damages bill over a trillion dollars.

Only a handful of authors were involved in the lawsuit, but it secured class action status, meaning any US author whose books had been illegally downloaded by Anthropic could be due a damages payment. The Chamber Of Progress also complains about the impact of class action copyright lawsuits in its new submission.

All of this together meant the Anthropic ruling was good news for the copyright industries, even though the judge had ruled AI training was fair use. The legal debates about the copyright obligations are far from resolved, not least because even the landmark cases are still at an early stage, given all initial rulings are likely to be appealed. 

However, legal uncertainty coupled with potential future statutory damages in class action lawsuits running into the billions if not trillions provides a strong incentive for AI companies to settle with relevant copyright owners and agree a licensing deal.

In the various music industry lawsuits filed against Suno and Udio, the major labels and various litigious artists have added allegations that both AI companies curated their training data by scraping audio off YouTube, in doing so violating rules in the Digital Millennium Copyright Act that prohibit the circumvention of YouTube’s technical protection measures.

Different statutory damages rules apply in that scenario, which the Chamber Of Progress describes in its filing. “Damages are potentially smaller, between $200 and $2500”, it explains, but that’s “per act of circumvention, not per work”. 

So if Suno or Udio scraped the same track from YouTube multiple times, damages would apply multiple times. Plus the DMCA allows damages to be tripled in cases involving repeat offenders. So we are still talking about potentially mega-damages.

According to Chamber Of Progress, the combination of statutory damages and class actions has “created a dysfunctional marketplace”, where licensing, litigation and the “arc of innovation” are “determined not by creative value or technological merit, but by justifiable fear of legal action”. 

This “coercive dynamic” is forcing AI companies into premature settlements which stops the world at large  from getting “judicial clarification” regarding the fair use dispute, which in turn “entrenches uncertainty”.

It then argues that the statutory damages system is not fit for purpose, because it was conceived “half a century ago in a regime meant for discrete acts” - ie when a potential infringer would infringe a relatively small number of works. Not AI companies that have potentially infringed millions of tracks. Applying that system to the AI sector, it argues, “amplifies risk in ways Congress could not have anticipated”.

To that end the Chamber Of Progress wants Trump to instruct his Department Of Justice to intervene in the various ongoing AI lawsuits by stressing, in court submissions, the AI ambitions of the US government and how the risk of mega-damages in copyright infringement cases is hindering those ambitions. And then to review the statutory damages framework and propose reforms to Congress. 

The copyright industries will be scathing of these arguments, stressing that the way for AI companies to remove the risk of trillion dollar damages is to negotiate upfront licensing deals. And while that may cost billions in licensing fees, AI companies with valuations in the hundreds of billions can afford to do that.

But the tech sector is likely to rally behind the Chamber Of Progress’s arguments. And not just AI companies. It was statutory damages that resulted in ISP Cox Communications being ordered to pay the majors a billion dollars in damages after being found liable for its users’ music piracy. Cox is fighting that judgment in the US Supreme Court right now. 

Great! You’ve successfully signed up.
Welcome back! You've successfully signed in.
You've successfully subscribed to CMU | the music business explained.
Your link has expired.
Success! Check your email for magic link to sign-in.
Success! Your billing info has been updated.
Your billing was not updated.
Privacy Policy