The Justice Department has put the US government behind OpenAI's fair-use defense in The New York Times copyright case. If you build AI models on web text, this is the clearest legal signal Washington has given you yet.
The Justice Department sided with OpenAI and Microsoft this week in the copyright lawsuit that could decide how every AI company in the country is allowed to train its models. In court papers filed Tuesday in Manhattan federal court, government lawyers argued that training large language models on copyrighted text can fall under fair use. The filing doesn't hedge. The United States has a strong interest in a court rejecting the argument that training LLMs on copyrighted texts violates copyright law, the department wrote, according to reporting from The Washington Post and WIRED.
That's a blunt sentence for a government brief.
It's also a first of its kind intervention. As WIRED reported, the Trump administration submitted a letter backing OpenAI's position after years of lawsuits from authors, publishers, music labels, artists and news outlets over what AI companies copied to build their systems. The Times sued OpenAI and Microsoft in December 2023, saying the companies used millions of its articles without permission to train products including ChatGPT and Microsoft Copilot. Judge Sidney Stein let the core copyright claims move forward in March 2025, while trimming other parts of the case. By April 2025, related OpenAI copyright cases had been centralized in the Southern District of New York for coordinated pretrial proceedings.
Why the DOJ got involved #
Its argument leans less on newspaper economics than on American AI power.
The New York Times filed a third amended complaint on June 25 directly targeting Microsoft's supercomputer platform, alleging it was purpose-built to scrape and replicate copyrighted content for AI training. The filing, arriving alongside a separate lawsuit from nearly 400 newspapers, sharpens liability exposure for every AI company relying on... - microsoft ai training lawsuit details - nyt copyright case against microsoft
Force AI developers to license every article, book and image they train on, and you hand the biggest technology companies an effective moat. That's the government's reasoning. Smaller AI startups without deep pockets for licensing deals would get frozen out. The money would flow to the publishers sitting on the biggest archives, not to individual creators. The Justice Department also argued that higher training costs could leave American AI companies at a disadvantage against overseas rivals.
That's the China argument, more or less. It's one OpenAI has made itself in prior statements, warning that tight limits on training data could weaken the US position in AI. You don't have to buy every part of that claim to see why it matters in court. Copyright law is being asked to carry a fight about labor, media revenue, national security and who gets to build the next layer of the internet.
If you're running an AI startup that trains on text scraped from the open web, this filing just gave you something unusual: an explicit government endorsement of the legal theory your business depends on.
The pushback is just as clear #
The Times isn't taking it quietly. Graham James, a spokesperson for the paper, told WIRED and The Washington Post that the administration is siding with "a handful of trillion-dollar AI companies" at the expense of American creators whose work they stole. Both AI companies and creators can thrive, he said, but AI companies need to pay fairly for the content that makes their products possible.
The Authors Guild was just as pointed. The group sued OpenAI in 2023 on behalf of writers including John Grisham, George R.R. Martin and Jodi Picoult, and its chief executive, Mary Rasenberger, told WIRED she was "extremely disappointed" by the DOJ letter. She called it "replete with faulty arguments" and a misunderstanding of fair use and copyright law.
Frankly, a statement of interest doesn't decide the case. Judge Stein isn't bound by the Justice Department's view, and the Times still gets to press its claims over copying and the market harm from those outputs. But a federal agency telling a judge, in writing, that OpenAI's theory carries the backing of the United States government is not nothing. It gives OpenAI's lawyers a citation they didn't have a week ago.
The timing is pointed too. Anthropic, Meta and other AI companies are fighting their own copyright cases, while several large media groups have signed licensing deals with AI firms as a hedge against losing this argument in court. The Associated Press struck an OpenAI deal in 2023. The Financial Times followed in 2024. Those deals don't settle the Times case, but they show the split in the market: some publishers are taking checks while others are asking a judge to say the copying was unlawful from the start.
OpenAI Plugs ChatGPT Into Epic's Health Records for 325 Million Patients OpenAI has integrated ChatGPT with Epic's electronic health record system, giving clinicians at UCSF Health read-only access to notes, labs, medications and specialist records for more than 325 million patients. The rollout adds role-based access, SSO, audit logs and BAA support for HIPAA compliance as OpenAI pushes deeper into hospital... - ChatGPT integration with Epic electronic health records system - how doctors access patient records through ChatGPT pilot
A DOJ win for OpenAI wouldn't bind every other case. It would still travel. Every AI company's lawyers would cite it, every publisher's lawyers would try to narrow it. And every founder training a model on public text would read it as permission to keep going until a court says otherwise.
Also read: GitLab stock jumps 20% as AI code governance demand outruns growth fears • Cloudflare Starts Blocking AI Crawlers by Default on Ad Pages This Week • Google Ships Gemini 3.8 Flash, Its Third Flash Model in Six Weeks