By Saxby Chambliss and Kent Conrad
China is racing to out-pace America in artificial intelligence (AI) development and deployment, and it is making serious progress. To stay ahead, America needs a national innovation strategy that strengthens all the pillars of our leadership, from data and infrastructure to talent and adoption. Thankfully, political leaders on both sides of the aisle are waking up to the intensity and importance of this race and many are implementing smart policies to ensure America maintains our edge.
Why America Needs A National Innovation Strategy
If we want to lead in AI, America needs a national innovation strategy grounded in four parts:
-
Policy Clarity: Freeze patchwork state regulations and preserve pro-growth legal frameworks, including supporting antitrust law that fosters growth and maintaining the current “fair use” copyright law to help catalyze innovation.
-
Infrastructure: Build the compute power, energy, and data systems needed to train and deploy AI at scale.
-
Talent: Invest in Science, Technology, Engineering, and Mathematics (STEM) education, upskilling, and skilled trades to propel innovation.
-
Adoption: Ensure our AI models are adopted at home and abroad so that American values, creativity, and innovation continue to underpin the global tech infrastructure, rather than China’s authoritarian vision.
What is Fair Use and Why It Matters
For nearly 50 years, fair use has allowed creators, educators, researchers, and entrepreneurs to use snippets of copyrighted material to produce transformative new inventions, and not just market substitutes for the original. In the early days of the internet, fair use was the legal bedrock that enabled search engines to index the web, web archives to preserve knowledge, and platforms to build global communities. The law has worked well to expand innovation and benefit the public, including by enabling researchers to take advantage of academic texts and developers to train early AI models on diverse, real-world content.
We didn’t strangle the internet in its infancy with impossible licensing requirements – and we shouldn’t make that mistake now with AI. Today’s large language models (LLMs) use vast public datasets to learn how language, code, and medicine work – not to reproduce them, but to understand and create new ideas and new solutions to our most pressing societal problems. Access to this data is essential to modern AI-powered tools that in turn allow small businesses to grow, doctors to discover treatments, and students to learn through personalized content. That’s exactly the kind of innovation Congress intended to protect with fair use. And it’s the kind of innovation we can’t choke off if we want America and our values to keep leading the world.
While We Debate a Tried and True Principle, China Charges Ahead
While some in the U.S. are challenging this this longstanding principle, China is using every speck of data possible to train its AI models and get a leg up in our fierce competition. That includes public data, government surveillance feeds, and personal information harvested at national scale.
In fact, China is pouring over $1.4 trillion into advanced technologies, including AI. It also uses hacking and spying to capture $500 billion in American tech and trade secrets each year, and is now using AI to export suppression, surveillance, and political control around the globe.
The stakes are high: Newsweek recently exposed how China is training AI to manage simulated cities to tightly control speech, behavior, and movement according to Communist Party values. If Chinese models become the global standard, we’ll see freedoms erode worldwide.
How Radical Changes to Current Fair Use Law Would Undercut American AI
Some want to force every AI developer to get explicit permission – and pay licensing fees – for every piece of content used in training models. This would break from precedent, kneecap innovation, bankrupt startups, and could hand China a lasting edge in global tech leadership.
Today, there are more than 30 active lawsuits against AI companies alleging that using copyrighted material to train models constitutes willful infringement. Some plaintiffs are asking courts for multibillion dollar fines that would effectively shut down dozens of U.S. AI firms overnight – an outcome that China would welcome with open arms.
If these lawsuits succeed – or if Congress radically rewrites the law – it will become nearly impossible for startups, universities, or even mid-sized firms to develop competitive AI tools. U.S. firms could move to more favorable foreign jurisdictions, driving jobs, talent, and capital out of the U.S. Such changes could also damage America’s open-source AI movement, one of our strongest checks against the global spread of Chinese technology.
The Stakes Are Too High to Get This Wrong
America’s tech leadership is under threat from Beijing to Brussels: China seeks AI dominance while the EU punishes American innovation through overregulation. Meantime our own lawmakers have introduced more than 1,000 AI-related bills in state legislatures this year alone.
Maintaining our current fair use framework supports American creativity and technological leadership. If we do away with it, we will unintentionally hobble our innovators and allow China to surge ahead, risk more dependence on foreign technology, have less freedom in the digital space, and face a global playing field redefined by authoritarian actors. Fair use is and has been a cornerstone of American innovation. If we want to win the AI future, we must ensure this critical component of copyright law remains intact.
###