Who Sets the Rules for Artificial Intelligence?

From Asimov’s fictional laws to real-world legislation in the Colombian Congress: who is truly in control when algorithms begin making decisions for us? Discover why voluntary ethics are no longer enough and how we are preparing to 'tame' AI without stifling innovation before the clock runs out."
— An article by the Ruta N Thought Center

Artificial Intelligence is advancing at a pace rarely seen in the history of technology. It can diagnose diseases, draft complex texts, optimize entire cities, and automate decision-making. But alongside these feats, an inevitable question arises: **Who ensures that all of this is used fairly, impartially, and safely?

This concern is far from new. In 1942, science fiction writer Isaac Asimov envisioned the famous "Three Laws of Robotics" in his short story *Runaround*. These rules—protecting humans, obeying orders, and self-preservation—were a hypothetical solution to a very modern problem: how to coexist with increasingly intelligent machines. Though born in fiction, they now serve as a powerful metaphor for today’s debate on technological governance.

Why Self-Regulation Isn’t Enough

AI development currently follows a set of voluntary ethical pillars: transparency, fairness, and explainability. However, international experience has shown that principles and self-regulation alone are insufficient. Cases of algorithmic bias, the spread of misinformation, and the misuse of personal data have proven that public trust in AI can erode rapidly.

Even tech leaders are calling for clear boundaries. At the *India AI Impact Summit*, OpenAI CEO Sam Altman warned that innovation without governance carries significant risks. The global conversation has shifted: the question is no longer *if* we should regulate AI, but *how* to do it without stifling innovation.

Colombia: Building on Established Foundations

Colombia is not starting from scratch. Over the last few years, the country has laid critical groundwork:

Ethical Framework for AI (2021).

National Policy on Artificial Intelligence – CONPES 4144 (2023).

Data Protection Authority Guidelines on automated systems (2024).

The core of the discussion has now moved to Congress, where six bills are being debated across four key levels. The goal is to move from abstract principles to mandatory rules:

1. A Comprehensive Risk-Based Framework: A proposal to classify AI systems by risk level, requiring impact assessments for high-risk applications and creating a specialized national authority.
2. Sector-Specific Regulations: Bills targeting specific uses—such as AI in citizen services or consular operations—emphasizing human oversight and data protection.
3. Protection of Vulnerable Populations: Specific rules for AI use involving children and adolescents, with a focus on psychosocial impact and digital equity.
4. Institutional Strengthening: The creation of a permanent legal commission within Congress dedicated to Artificial Intelligence.

The Path Forward

Colombia needs to regulate AI, but the challenge lies in doing so without fragmenting the legal system. Effective regulation isn't about multiplying the number of laws; it’s about building a coherent architecture that is proportional to risk and technically informed.

Regulating AI is not a bureaucratic whim. It is a fundamental requirement for the technology to be socially legitimate and sustainable.

Asimov intuited this through fiction: powerful technologies need understandable boundaries to coexist with humanity. Today, instead of imaginary laws for robots, we need real regulatory frameworks for algorithmic systems.

Colombia is at a decisive turning point. We have the opportunity to build "smart regulation" that protects people while driving innovation. The question is no longer whether we should set rules for AI.

The question is whether we can do it well—and do it in time.

¿Te gustó este contenido? Conoce otros artículos de Centro de Pensamiento Ruta N como:

Ruta N: 4 claves para entender cómo el Foro Davos 2026 puede impactar el ecosistema CTi de Medellín

"Beby, te vo’a dar una vuelta por..." Medellín 🐰 tecnología, turismo y uso responsables de las aplicaciones

Bibliografía  

Congreso de la República de Colombia. (2025a). Proyecto de Ley 043 de 2025 Senado / 324 de 2025 Cámara: Por medio del cual se regula la inteligencia artificial en Colombia. https://www.camara.gov.co/wp-content/uploads/2025/12/proyectosley/publicaciones/proyecto-35981/PPDSC-PL-043-25S-324-25C-INTELIGENCIAARTIFICIAL-SC.pdf

Congreso de la República de Colombia. (2025b). Proyecto de Ley 417 de 2025 Cámara: Regula e implementa el uso de IA en la gestión de PQRSD en entidades públicas del Estado colombiano, y dicta otras disposiciones.

Congreso de la República de Colombia. (2025c). Proyecto de Ley 384 de 2025 Cámara: Regula el acceso y uso de plataformas digitales de IA en niños, niñas y adolescentes.  

Congreso de la República de Colombia. (2025d). Proyecto de Ley 098 de 2025 Cámara: Establece un marco regulatorio para el desarrollo y uso de IA con enfoque psicosocial y equidad digital.https://www.camara.gov.co/wp-content/uploads/2025/10/proyectosley/documentos/proyecto33950/p_l_098_2025sc_marco_regulatorio_de_inteligencia_artificial-f6c87c61.pdf

El Colombiano. (2026). Altman pide regulación urgente de la inteligencia artificial en cumbre mundial en India.

Asimov, I. (1950). Roundabout. En I, Robot. Gnome Press.