OpenAI and Google's dark new campaign to dismantle artists' protections

OpenAI and Google have cozied up to Trump's imperialist approach to AI for a shot at tearing down artists' copyright protections.

Greetings all,

So I spent much of the week traveling, and the rest of it sick. But I’m back from a long and good and productive SXSW, where I did a featured talk with the great crypto skeptic Molly White, a chat at the LightHouse about generative AI and labor, and a talk with the 404 Media folks about AI slop at the Fediverse House. If video for any of those talks surfaces, I’ll share them. I also wrote a big and probably overlong response to all the AGI talk going on right now, but given the exhaustion and the brian fog I figured I’d wait to give the thing an edit; look for that early next week.

Thanks as always to everyone who supports this work—you alone make this writing, reporting, and research possible. If you’d like to help ensure I can keep at it, please consider subscribing for $6 a month; a cheap beer or a higher end cup of coffee. Thanks again.

One of the biggest hurdles to generative AI’s dominance—and one of working artists’, writers’, and creators’ only tools to protect their livelihoods from its creep—is copyright law. And the two biggest AI companies in Silicon Valley just took a big, well-calculated swing aimed directly at breaking it down this week.

The Trump administration is currently soliciting public comment on its AI Action Plan, whose aim is to “sustain and enhance America’s global AI dominance.” OpenAI and Google cannily took the opportunity to try to get the Trump admin to help them remove one of the biggest albatrosses from their necks.

Image: elizal, via Flickr. CC BY SA 2.0

Along with proposing that Trump adopt a hands-off regulatory approach and implement policy that promotes the export of American AI systems, OpenAI proposed that the administration officially enshrine AI companies’ practice of using copyrighted works in their training data as fair use. Right now, the legality of that practice is being hotly debated in the courts, and artists, authors, and journalists have sued the AI firms for ingesting their data, using it to train AI models, and producing commercial output with it—all without the rights holders’ consent or compensation.

Last month, Thomson Reuters won the first major copyright victory against AI companies—though notably not generative AI firms, and the lawsuit was filed before the rise of ChatGPT—on the grounds that the firm violated fair use laws. The output of the AI models competes with the materials it was trained on, the judge found—a similar contention that is made by claimants like the New York Times in its case against OpenAI.

Meanwhile, an ongoing class action lawsuit brought by artists including Karla Ortiz—which AI firms had pressed to get dismissed—was allowed to go to discovery. The AI firms certainly hoped the case wouldn’t make it that far, and as of now they face a real possibility that their practices violate current copyright law.

Last year, the artists—those organizing against the use of AI in their workplaces, and those waging the class action lawsuits against the AI giants—were gaining ground. Now all that is at risk of being wiped away.

Which is why OpenAI is going scorched Earth on the matter, and appealing to Trump and Vance’s freshly imperialist approach to AI—touting the need to secure US AI dominance at all costs, and casting China as a foe; pointing to the widely discussed DeepSeek in the process.

In its lengthy proposal document, OpenAI argues that, “if the [People’s Republic of China]’s developers have unfettered access to data and American companies are left without fair use access, the race for AI is effectively over. America loses.”

OpenAI is certainly reading the room here—this is exactly the kind of rhetoric the Trump administration has been embracing around AI. A month ago, in his first foreign policy speech as vice president, JD Vance declared that an age of American dominance in AI was beginning, that safety talk would no longer be a concern, and that nations had a choice; whether to partner with the US, or with China.

And here’s the language in the Trump admin’s Action plan:

The AI Action Plan will define priority policy actions to enhance America’s position as an AI powerhouse and prevent unnecessarily burdensome requirements from hindering private sector innovation. With the right governmental policies, continued U.S. AI leadership will promote human flourishing, economic competitiveness, and national security.

OpenAI is using that framework as a means of advancing its case for killing creators’ copyright protections—it’s a shrewd and disturbing move, and one that’s designed to solicit the approval of the Trump admin’s hawkish approach to AI.

Google, for its part, isn’t quite as aggressive; it only mentions China twice in its own lengthy proposal document, but does make sure to name-check Vance and his Paris speech. Its first ask is for energy policy that ensures data centers have ample supply, but Google too seeks to end AI’s copyright complications.

From its proposal:

Balanced copyright rules, such as fair use and text-and-data mining exceptions, have been critical to enabling AI systems to learn from prior knowledge and publicly available data, unlocking scientific and social advances. These exceptions allow for the use of copyrighted, publicly available material for AI training without significantly impacting rights holders and avoid highly unpredictable, imbalanced, and lengthy negotiations with data holders during model development or scientific experimentation.

Google wants an exception granted to AI companies, like itself, in copyright law, in large part to avoid “unpredictable… and lengthy negotiations” that might, in other words, require it to compensate creators for the work it uses to train its models.

The difference in the tenor of both documents is interesting as well, as a reflection on the two tech companies’ stature. Google, as a search monopoly with dominion over vast stretches of web infrastructure, is dull and workmanlike in its proposal; clearly it would love to do away with a copyright rule that proves a minor thorn in its side. But life would go on if the Trump camp was unpersuaded, and the courts ruled the other way. For OpenAI, whose business model is already volatile and uncertain, a ruling that its practices violate fair use is more of an existential-level threat; read both these proposal documents back to back, and the difference in gravity is clear.

But either way, this is dark stuff: A $160 billion AI company and a $2 trillion tech giant invoking the need to maintain technological—and implicitly, military—superiority over China, the former casting it as a foe, to justify eroding artists’ already meager rights and protections. Perhaps AI really has entered its empire era.