The debate over regulating artificial intelligence usually focuses on two competing visions. In Europe, lawmakers are writing detailed rules that govern how AI can be developed and used. In the United States, policymakers are taking a lighter touch, allowing companies, investors and consumers to shape the technology’s future.

But a new analysis from students at the University of Florida identifies a third force quietly shaping the future of AI in America: the courts.

As AI spreads faster than any previous technology, judges and juries are being asked to resolve disputes. In doing so, they are not simply applying existing laws—they are, case by case, defining what responsible AI use looks like. The result is a distinctly American form of AI governance: one built through the give and take of negotiations and legal processes rather than legislation.

So far, courts have mostly resisted treating AI as something fundamentally new. Instead, they have folded AI into existing legal doctrines, focusing on the humans and institutions behind the technology.

Consider employment discrimination. In EEOC v. iTutorGroup, an AI-driven hiring system screened out older applicants. The case ended in a settlement, but its message was clear: Employers remain responsible. The technology does not shift accountability.

This pattern repeats across much of AI litigation. Courts are less concerned with the machines than with whether their use allows people to avoid responsibility.

But that approach becomes harder to sustain as cases move into less familiar terrain. In defamation disputes, for example, courts are beginning to grapple with what it means for an AI system to “say” something false. In Walters v. OpenAI, a Georgia court rejected a claim based on erroneous chatbot output, emphasizing context, disclaimers, and the lack of actual malice. The ruling suggests users should exercise caution when using AI that hallucinates.

Other cases are pushing even further. In Garcia v. Character Technologies, Inc., a federal court allowed negligence claims to proceed against a chatbot developer in a tragic case involving a teenager’s suicide. The decision did not address whether chatbot outputs are protected speech, but it made clear that courts will examine AI systems as products that may carry duties of care.

The most consequential disputes may be in copyright law. Here, courts are drawing lines that could determine the future of the AI industry. Cases such as Andersen v. Stability AI and The New York Times Co. v. OpenAI challenge whether companies can train AI systems on copyrighted works without permission. These cases, although still open, highlight a striking tension. Humans are free to learn from copyrighted material and use the knowledge gained. AI systems may face stricter limits. How this is resolved will shape not only the economics of AI but also incentives for creativity and innovation.

Geography is also playing a role. Many of these cases are concentrated in California and New York—home to the nation’s technology and media industries. As a result, a relatively small number of courts are exerting outsized influence over AI.

This judicial activity is occurring against a backdrop of limited legislation. Every state has considered AI-related legislation, but only a few bills have been enacted. Congress has yet to act. In this environment, courts are resolving issues, relying on common-law reasoning to extend accepted principles to new technologies.

This approach has advantages. Common law evolves incrementally, allowing rules to adapt as technologies and markets change. It can accommodate uncertainty and avoid locking in premature assumptions about how AI will develop.

But it also has costs. Litigation is slow and expensive. It places the burden of shaping legal rules on the parties involved in individual disputes. And it creates uncertainty, as businesses and consumers must navigate a patchwork of emerging precedents.

By contrast, legislative approaches—such as those adopted in Europe—can provide some clarity but risk creating rigid requirements that are quickly outdated.

The US is, for now, pursuing a different path, allowing rules to emerge through disputes over hiring, speech, copyright, and liability. Judges may not be setting out to regulate AI. Yet in resolving these cases, they are defining the boundaries. This is a system that reflects a familiar American instinct: to let practices develop through experience rather than design.