Artificial intelligence is growing exponentially fast, and the laws meant to govern it, in my opinion, are trailing behind. The problem for me is simple: AI does not respect borders. A system built in one country can instantly reach people in another, leaving regulators struggling with what to do.

So, what happens when technology moves faster than the law? In short, things get complicated. Answers to this question can be found in the current regulatory landscape.

In 2024, the European Union released the world’s first comprehensive AI regulatory framework, classifying AI by risk level. Unacceptable-risk systems (such as biometric categorization systems or social scoring) are prohibited, high-risk systems face strict compliance standards regardless of where their developers are based and limited-risk systems like chatbots simply must disclose their AI nature to users.

General-purpose models like ChatGPT must document their systems and follow copyright law. Any model posing systemic risk must undergo safety testing and incident reporting. Minimal-risk AI, like video games, is left unregulated entirely.

In contrast, the U.S. has taken a fragmented, largely voluntary approach to AI regulation. Instead of comprehensive federal legislation governing AI, the government has encouraged industry self-regulation. This leaves federal agencies like the Federal Trade Commission (FTC) and the U.S. Food and Drug Administration (FDA) to apply existing laws to AI on a case-by-case basis, creating discrepancies.

Executive orders have addressed some divides, but without Congress passing conclusive legislation, enforcement remains inconsistent. This has forced regulation down to the state level. More than 1,000 AI-related bills were introduced across nearly every state in 2024 and 2025, leading to a miscellany of conflicting rules that varied drastically by region.

For businesses operating across the nation, compliance becomes tangled. For users, protections depend largely on the state they live in, which is a stark contrast to the EU’s uniform legislation.

The key jurisdictional problem I see is this: AI systems are built in one country, shaped on data from another and used everywhere. When something goes wrong, which country’s legal jurisdiction applies? When viewed this way, unanimous enforcement across borders seems practically impossible.

In my opinion, this discussion matters hugely from an ethical utilitarian standpoint. AI affects everyone globally, and its risks to the most pressing issues – privacy, employment and human rights – fall hardest on those most susceptible to vulnerability. I believe this alone makes a strong case for interrelated global regulation, but getting there is the tough part.

Many countries are reluctant to give up regulatory control. The U.S. and China are going head-to-head in an AI race with little incentive to agree on strict shared rules. Laws struggle to keep up with the constantly changing technology.

Many lawmakers lack the technical knowledge to write effective legislation, and there isn’t even a universally agreed-upon definition of what AI actually is, making any solid international framework difficult to implement widely.

So what could actually work? I believe the most rational and beneficial path may be agreeing on a set of minimum ethical standards – focused on transparency, accountability and civil liberties – while giving individual countries flexibility in how they implement them.

This solution may not be perfect, but it may bring some unification amongst jurisdictions. When technology moves faster than the law, a dangerous divide opens up between powerful tools and the lack of accountability. With the rampant, explosive growth of AI, time is of the essence.

As we’ve seen, by the time legislation passes, it’s responding to yesterday’s AI. The EU’s AI Act took years to develop, and many argue it’s already falling behind. In my opinion, the U.S. remains fragmented, and globally, no binding framework exists.

Questions surrounding global AI regulation are not just technical; they are deeply complex. One thing I am confident of is this: effective and ethical AI regulation across jurisdictions is necessary. The challenge is finding the common ground to make it happen before the divide grows too deep to close.