President Donald Trump is siding with technology firms in the ongoing dispute over AI copyright. On Wednesday, he unveiled his AI Action Plan during a speech at the AI Summit, organized by the All-In podcast and the Hill and Valley Forum. Trump pushed for a “common-sense application” of AI and intellectual property regulations. “You shouldn’t be expected to run a successful AI program when you’re required to pay for every article, book, or anything else that you’ve engaged with,” he remarked. “When an individual reads a book or an article, they’ve acquired significant knowledge. That doesn’t imply you’re infringing copyright laws or must negotiate with every content creator.” He further commented, “China’s not doing it.”
The sentiment resonates with the stance of leading tech firms such as Meta and OpenAI regarding AI model development and copyright violations. A fierce legal conflict is in progress, with copyright owners contesting AI enterprises. Authors, musicians, and artists have filed lawsuits against OpenAI, Meta, Anthropic, and others, claiming that training their creative works amounts to copyright infringement. Disney, for instance, described the AI image generator Midjourney as “a bottomless pit of plagiarism” in its lawsuit. AI firms maintain they are shielded by the fair use doctrine, which allows unlicensed use of copyrighted materials under certain conditions.
In March, Meta urged the Administration to clarify that utilizing publicly accessible data to train models is fair use. OpenAI likewise contended that the application of the fair use doctrine to AI is vital for American competitiveness and national security. Companies like OpenAI argue they require authorization to use copyrighted works without consent to remain competitive against China in the AI sphere.
As part of its AI Action Plan, the White House released executive orders regarding AI, which Trump endorsed following his speech. Although these orders did not specifically address copyright concerns or fair use, Trump’s statements imply that the administration is in favor of large tech on this issue. The action plan also suggests reducing regulations on tech firms. “To uphold global leadership in AI, America’s private sector must be free from bureaucratic obstacles,” the plan asserts.
Trump’s assertions about not needing to pay for acquiring knowledge reflect a prevalent argument. Tech companies argue their AI models primarily learn from data, thereby not falling under copyright laws. However, numerous copyright attorneys disagree. Justin Hughes, a law professor at Loyola Law School, noted that while the challenge of compensating for every “single article, book, or anything else” is valid, AI companies ought to license vast collections of copyrighted content, come to agreements with copyright holders, and pay damages for unauthorized copying. Hughes also remarked that individuals invest in learning and must purchase books to attain knowledge.
Several AI copyright lawsuits are moving through the judicial system. The blog ChatGPT Is Eating the World reports 47 ongoing cases. Recently, Anthropic and Meta achieved favorable fair use rulings, but the judge in the Meta case clarified that the ruling doesn’t indicate that Meta’s use of copyrighted material to train its models is legal and suggested it might be unlawful to replicate copyrighted works for AI training without prior consent.
It remains unclear how Trump’s comments on copyright will manifest in actions. He seems misaligned with the U.S. Copyright Office’s perspective on AI training and fair use, as indicated in a pre-publication draft of its AI Copyright Report. The report concluded that fair use likely does not offer a blanket exemption for AI companies utilizing copyrighted materials. In Congress, Senators Josh Hawley and Richard Blumenthal recently put forth a bipartisan bill to prohibit AI companies from training on copyrighted works and to empower individuals to sue tech firms for using their personal data in model training.
Hughes speculated that Trump’s view might shift if he became aware of AI companies training on episodes of The Apprentice without compensation.
Disclosure: Ziff Davis, the parent firm of Mashable, initiated a lawsuit against OpenAI in April, claiming copyright infringement in the training and operation of its AI systems.