As AI companies grapple with how to securely scale their models without alienating a public that's grown weary of data center creep, job loss, and the overall flattening of language, President Donald Trump has a message for frontier labs: Don't take your foot off the gas. His reluctance to regulate AI, however, could be bad news for publishers trying to get these companies to pay up.
“We could put guardrails. We can do this and that," Trump said this weekend. "But I think you have a lot of very negative forces that are...bringing up things that won’t happen."
He later went full throttle on Truth Social, posting a series of diatribes that likened concerns about AI killing us all to the investigation into his first presidential campaign's ties to Russia and climate change. "I am the Hoax Buster, and I’m right now breaking another Hoax," he wrote.
Later, he pushed AI companies to ignore calls for a slowdown in development. "Concerning AI, when, in the History of Business, did anyone see the Leaders of an Industry call for Regulation that, if strongly implemented, will drive them into oblivion and bankruptcy?" he wrote. "President Xi, of China, just announced that China will be doing absolutely nothing to stand in the way of AI, or its future. Google has recently stated that they want to build a massive Plant in Finland, all because they are finding permitting too difficult in the United States. I am not happy about this, and want them to change their thinking. AI, and Data Centers, will be the Greatest Economic Development Engine in History — Bigger than Oil, Gold, Diamonds, or even the Internet. It will not be stopped by brilliantly run Destructive Forces during the Term of President DONALD J. TRUMP!"
His stance doesn't bode well for publishers suing OpenAI, Google, Microsoft, and Anthropic for training their large language models (LLMs) on journalists' content without permission. If your focus is on beating China in the AI race at all costs, slowing down the progress of US companies by backing huge copyright payouts is probably not top of mind, particularly for an administration that has had a particularly testy relationship with the press.
News publishers argue that their content has been scraped and stolen by LLMs to generate answers and knowledge graphs without compensation. In 2024, a few months after it sued OpenAI and Microsoft for copyright infringement, The New York Times said it had found substantial evidence that OpenAI stole data from various news publishers to help build its tools and generate answers to queries.
AI companies have signed a handful of licensing deals with publishers, but it's not applied evenly. And the US government is siding with OpenAI. Earlier this month, the Department of Justice submitted a Statement of Interest to a New York court arguing that training AI models on copyrighted material does not violate copyright law, citing fair use. "Constraining LLM development under a misunderstanding of fair use doctrine would thwart...creative and scientific progress while hindering American prosperity and economic mobility," the DOJ says.
'Where Exactly Is the Difficulty?' #
That argument caught the eye of Vivek Shah, CEO of Ziff Davis, PCMag's parent company. Ziff sued OpenAI last year after it "relentlessly reproduced exact copies and created derivatives" of articles from Ziff Davis publications, including PCMag. Now, "the government concedes, in a footnote, that it 'takes no position on whether a licensing regime would be financially or logistically feasible' and that 'both mainstream and independent publishers could enter (and have entered) into licensing agreements," Shah writes for Fortune. "So, where exactly is the difficulty? What onerous obligation are we placing on frontier labs that would slow their progress?
Shah notes that OpenAI expects to spend about $750 billion on compute by 2030, and points to the $20 billion per year in royalties and licensing paid to the US music industry.
"If similar amounts were paid to news publishers, it would represent a rounding error to LLM companies’ overall expenses, but be meaningful for content producers of all sizes," he writes. "A sensible fee structure creates a flywheel of quality inputs and quality outputs, benefiting the AI consumer and the public good."
What's the Problem With AI Training? #
Search engines like Google have long scraped web content and directed people to content through the familiar blue links. It's been a lifeline for publishers who were forced to abandon print and go all-digital, even though Google's own ad network forced that pivot.
Now, however, AI chatbots use that information wholesale or rewrite it. They include links to the original sources, but few people click through because they've learned what they need to know through AI summaries. Traffic to publishers plummets, as does revenue.
In this landscape, how can media companies make enough money to stay in business? One idea is that AI firms pay fair content licensing fees for the use of information. That would allow journalists to keep working, generate new information for AI companies to cite, and maintain their status quo.
Courtney C. Radsch, a Brookings Institution fellow, argues that AI platforms are the “new tollbooths” for media. But failing to fairly compensate media outlets will mean fewer of them exist, resulting in less content for AIs to scrape and train on, thereby making them less useful to users. This concept is tied to the dead internet theory, which posits an eventual reality in which AI bots become the primary users online, regurgitating the same information over and over without further input.
“AI systems depend on a continuous supply of high-quality human content to remain useful," Radsch says. "Degrade that supply—by destroying the economic conditions under which it’s produced—and you degrade the AI itself.”
AI firms may continue to sign licensing deals with news publishers; last year, The Washington Post partnered with OpenAI, and the Associated Press teamed up with Microsoft. However, smaller media outlets and publishers may not be able to negotiate large deals, so it will likely remain an uphill battle without a landmark legal ruling or substantial action from Congress. (Don't hold your breath on the latter.) In the interim, this may degrade your internet experience, as publications disappear and AI turns to less reputable sources.
As Shah says, "Without a licensing framework, publishers will increasingly erect barriers and paywalls to restrict their own content and extract those fees. This imposes substantial costs to consumers, limits access to free information and content on the open web, and reduces pathways to earnings for writers and content producers."