Regulating Artificial Intelligence: Who Decides the Rules of the Game?

Who writes the rules for artificial intelligence? Explore who decides, how AI regulation works, and what the future holds for AI governance.

When Technology Outpaces the Law

Today, we use artificial intelligence almost without noticing. It suggests what to watch, how to get around, what to buy. But have we ever asked ourselves: who controls all this? Who decides what an algorithm can and cannot do? Technologies move fast, laws much less so.

The issue of regulating artificial intelligence has become crucial. From big tech to governments, everyone is trying to establish a framework of rules. But how do you regulate something that evolves every day? And who has the right to do it?

What It Means to Regulate Artificial Intelligence

Regulating AI means defining limits, criteria, and responsibilities for those who develop, use, or are affected by intelligent systems. The goal is twofold: maximize the benefits of AI and reduce the risks for citizens, businesses, and society.

Regulations can cover various aspects: algorithm transparency, privacy protection, the absence of discrimination, liability in case of error. In practice, it's about preventing AI from becoming a "lawless space," like a digital no man's land.

Who Decides the Rules?

The rules for artificial intelligence are not written by a single authority. On the contrary, they are decided by a complex interplay of actors:

– The European Union, with the AI Act, the first global attempt to create comprehensive AI legislation.
– National governments, with different approaches: from China's rigid stance to the more self-regulated approach of the United States.
– Big Tech, which often set de facto rules through the massive use of their own platforms.
– International organizations, such as the UN, OECD, and the Council of Europe, which promote ethical guidelines.
– Civil society, academics, activists, and citizens, who demand more transparency and rights.

In our article The Ethics of Artificial Intelligence: Why It Concerns Us All we have already addressed the issue of responsibility. But today we go further: who really has the power to impose global rules?

Ethics, Power, and Artificial Intelligence

The problem is not only legal. It is also political and cultural. Regulating means deciding who can do what, who is protected, and who can profit. It means addressing the issue of algorithmic power, as explored in AI and Surveillance: Who Protects Us from Whom?.

An example? Predictive algorithms used in justice or credit. Who guarantees they are not discriminatory? Who can access their code? And if they make mistakes, who pays?

Without clear rules, artificial intelligence risks entrenching inequalities, amplifying biases, and undermining fundamental rights.

Practical Examples and Real Cases

In Europe, the AI Act regulation provides for a classification of AI systems based on risk: unacceptable (e.g., China-style social scoring), high (e.g., AI for recruitment or justice), medium, and low. High-risk systems will be subject to obligations of transparency, traceability, and human oversight.

In the USA, however, regulation is still fragmented. Some states, like California, have advanced privacy laws (like the CCPA), but a unified federal regulation is lacking.

Globally, the OECD has drafted guiding principles for responsible AI, while UNESCO has published an ethical framework for artificial intelligence.
👉 OECD AI Principles
👉 UNESCO Recommendation on the Ethics of Artificial Intelligence

These documents are voluntary, but they influence public policies.

Frequently Asked Questions (FAQ)

Does regulating AI slow down innovation?

Not necessarily. Good regulation can stimulate safe and responsible innovation, creating trust among citizens and markets.

Can AI be dangerous without rules?

Yes. Without adequate norms, AI can amplify biases, violate privacy, or be used to manipulate opinions, as happens with automated disinformation.

Can we have a global AI law?

It's difficult, but increasingly necessary. Today, the risk is that each country goes its own way, with incompatible standards. International cooperation will be fundamental.

Conclusion: Rules to Avoid Losing Control

Artificial intelligence is not neutral. Behind every algorithm, there are choices, interests, and consequences. Regulating it does not mean blocking it, but guiding it. And to do this, constant dialogue between governments, companies, and citizens is needed.

The questions are many: who oversees the algorithms? How are fundamental rights protected? Can we build a digital democracy where AI serves people?

The answer is not simple. But the time to ask these questions is now.

If you are also interested in the legal implications, you can read our in-depth analysis Digital Justice: Artificial Intelligence in the Dock?