Here's my infographic depicting Trump's possible open source AI model ban that will "keep America dumb" while China's open source models will surge ahead and empower the rest of the world with advanced machine intelligence.

Trump should call it "MADA."
Make America Dumb Again.

The Trump administration wants to ban access to open source AI for the exact same reason the FDA bans access to natural remedies: Because they are both protecting an inferior but far more lucrative domestic industry that depends on monopoly enforcement to exist.

Trump wants to force you to use inferior U.S. AI models that cost far more money but fail to deliver on intelligence and performance (often due to ridiculous guardrails), while the FDA wants you to depend entirely on expensive pharmaceuticals so that you never find out that your health issues can be prevented or cured for a fraction of the cost using natural remedies like herbs, essential oils, nutritional therapies or even just light therapy.

HUGE RULING says that training AI models on copyrighted content is "Fair Use" under U.S. law.

The Case: Bartz v. Anthropic

The lawsuit was filed in August 2024 by three authors: Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson, alleging Anthropic used millions of pirated books to train its Claude AI models.

What the Judge Ruled:

Training on copyrighted books = fair use

Judge Alsup ruled that Anthropic's use of copyrighted books to train its large language models was "quintessentially transformative" and "exceedingly transformative" -- protected as fair use under U.S. copyright law. He compared it to a human reading books to learn to write: "Like any reader aspiring to be a writer, Anthropic's LLMs trained upon works not to race ahead and replicate or supplant them -- but to turn a hard corner and create something different." He also found that training did not displace demand for the original books.
(Sources: Reddit, and Kimi K3 research agents)

This is exactly what I have personally argued for several years now. Reading books and learning from them is NOT copyright infringement. Distilling knowledge from books is how every university education works, obviously.

Anthropic was found, however, to be infringing on authors' rights by maintaining a library of "pirated" books in a commercial context for an unknown future use which seemed focused on profit. This is what the settlement actually addressed, NOT the use of such books for training LLMs.

In other words, if Anthropic had pirated millions of books, then used them to train their models, then DELETED the book library, the judge would have said that's all 100% Fair Use and there would have been no settlement at all.

And if that library of books had been maintained in a non-profit context for the purpose of preserving knowledge without commercial applications, there would have been no case at all, because even that would have clearly qualified as fair use. I'm not a lawyer, but that's how I see it. Look up the details of the case yourself if you want to get into the weeds on this.