Publishers Demand Fair Pay as AI Copyright Battle Reaches Critical Phase With DOJ Backing OpenAI

Publishers Challenge DOJ's Fair Use Position in AI Copyright Cases

The battle over AI copyright has reached a pivotal moment as copyright infringement lawsuits filed by major news organizations against OpenAI advance toward key rulings1

1

. Publishers including The New York Times and Ziff Davis are contesting whether OpenAI's use of published content to train AI models like ChatGPT constitutes fair use under US copyright law1

1

. The Department of Justice intervened with a Statement of Interest arguing that AI training on copyrighted material should be classified as fair use because it is "exceedingly transformative"2

2

. This unusual move places the government's weight behind AI companies in disputes with private parties, despite not being directly involved in the litigation2

2

.

The Economics of AI Training on Copyrighted Material

Ziff Davis CEO Vivek Shah pushed back forcefully against claims that licensing fees for AI content would burden AI firms. Writing in Fortune, Shah compared the $20 billion per year in royalties and licensing paid to the US music industry against the $750 billion that OpenAI expects to spend on computing infrastructure by 20301

1

. "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," Shah argued3

3

. Shah contends that sensible licensing fees would create a "flywheel of quality inputs and quality outputs" benefiting both AI consumers and the public good1

1

.

Trump Administration Prioritizes AI Development Over Publisher Concerns

The Trump administration has made clear its position on AI regulation, with President Trump stating over the weekend that concerns about AI development represent "negative forces" bringing up problems "that won't happen"3

3

. In a series of Truth Social posts, Trump urged AI companies to ignore calls for slowdowns, comparing AI safety concerns to "hoaxes" and emphasizing competition with China3

3

. "President Xi, of China, just announced that China will be doing absolutely nothing to stand in the way of AI," Trump wrote, framing AI and data centers as "the Greatest Economic Development Engine in History"3

3

. This stance suggests the administration will prioritize beating China in the AI race over backing copyright payouts for publishers vs AI companies3

3

.

OpenAI Defends Scraping Practices as Implicitly Licensed

OpenAI argued in its court filing that its use of data scraped from the internet was "highly transformative" and caused "no cognizable harm to the market"1

1

. The company claimed access to publishers' content was implicitly licensed because publishers didn't block AI scrapers using robots.txt or other mechanisms at the time1

1

. "For decades, search engines like Google have relied on scraping or crawling the Internet," OpenAI stated, arguing this copying has been understood as implicitly licensed unless website owners opt out1

1

. Many publishers now block bots from OpenAI and other AI firms, though Shah warns these barriers will grow more onerous, limiting access to information1

1

.

Publishers Document Market Harm From AI-Generated Content

In their joint brief, news publishers argued that ChatGPT's use of their content has undermined web traffic and flooded the market with AI-generated content1

1

. "One effect of ChatGPT's release in November 2022 was to spark other companies like Google to fast-track their own GenAI products, including, most significantly, Google's AI Overviews," the publishers stated1

1

. Sites have lost almost 40% of their search engine traffic on average due to AI Overviews, contradicting the DOJ's rejection that AI-generated works cause market harm2

2

. The New York Times found substantial evidence that OpenAI stole data from various news publishers to generate answers to queries3

3

.

The Scope of Intellectual Property Ingestion by AI Companies

AI companies have systematically ingested copyrighted material on a massive scale. Anthropic and other large language models developers are reported to be buying physical books by the millions, cutting them apart, scanning them, and disposing of the remains2

2

. Many creators discover their work in AI training corpora without ever being asked for permission or receiving compensation2

2

. This bulk ingestion of intellectual property has sparked multiple copyright infringement lawsuits from companies including Ziff Davis, which sued OpenAI after it "relentlessly reproduced exact copies and created derivatives" of articles from its publications3

3

.

What Comes Next for Publishers and AI Regulation

As the lengthy discovery process wraps up, courts will rule on key questions before sending individual cases back to their respective districts for potential trials1

1

. The DOJ OpenAI copyright stance and Trump administration's push for unrestricted development suggest publishers face an uphill battle. While some AI companies have signed licensing deals with publishers, these agreements aren't applied evenly3

3

. The outcome will determine whether national security concerns and competition with China override intellectual property rights, or whether courts will require AI companies to compensate creators whose work trains their models. Watch for rulings on fair use definitions and whether implicit licensing arguments hold up against publishers' documented market harm.