Colombia’s Ministry of Science, Technology and Innovation and the Ministry of Information and Communications Technologies (MinTIC) have formally submitted a draft bill to regulate artificial intelligence systems in the country. The bill (442-2025) seeks to regulate artificial intelligence in Colombia “to ensure its ethical and responsible development”.
A key aspect of the bill is the intention from the central government to establish a clear and robust legal framework to ensure the ethical, responsible, and rights-based development and use of AI technologies in Colombia.
Key takeaways
The draft bill, inspired by international standards from the OECD, UNESCO, and the European Union, tries to establish key principles to boost AI development and investment in the country. As such, one of its main aspects is the classification of AI risks, data governance, innovation incentives, fundamental rights, and cross-border cooperation. The authorities recognise the transformational potential of AI technologies and reinforce their interest in promoting economic development leveraged on AI innovations.
This draft bill, once approved, will apply to all natural (individuals) and legal persons (organisations), whether public or private, who create, develop, implement, commercialise, import, represent, distribute, or use AI systems within Colombian territory, or who, although located abroad, are subject to Colombian law. They are referred to as “AI Responsible Parties” and can be understood as potential regulated subjects.
Risk-based classification
One of the bill’s main components is the risk-based classification of AI systems, strongly inspired by international references such as the EU’s AI Act, UNESCO recommendations, and industry recommendations. There are four risk levels that could potentially guide secondary regulation, if needed.
- Unacceptable Risk: Prohibited uses (e.g., subliminal manipulation, social scoring).
- High Risk: Subject to strict oversight (e.g., in health, finance, education, justice).
- Limited Risk: Requires transparency measures.
- Minimal Risk: General use, with minimal oversight.
Why it matters
This legislative initiative adds to Colombia’s persistent AI efforts. For stakeholders—especially AI developers and companies, multinationals, and civil society—the bill brings both compliance obligations and opportunities to shape a high-standard, innovation-friendly ecosystem. The explicit international alignment may also facilitate cross-border digital trade and regulatory harmonisation. Consistent regulatory frameworks across the region allow for economies of scale and regional investment.
Next steps
The bill will now begin its congressional review, with a little more than a year until Congress (and President) elections take place in Q2-2026. If approved, it will establish a precedent in the region for rights-based, risk-conscious AI regulation. Nevertheless, with this one, there are 23 laws that address the regulation of AI from different perspectives. 11 of them are specifically addressing AI regulation, whilst the remaining 12 consider indirect regulations. None of them have been approved and it is yet to see if this initiative from the Colombian authorities will change the course of previous bills.
Companies operating or offering AI-based services in Colombia should assess their AI systems for compliance readiness and prepare for new documentation, transparency, and impact assessment requirements considering the regulatory intentions from the authorities.
As the leading tech policy firm, Access Partnership can guide you through this transformative period. Our expertise in regulatory advocacy and market strategies ensures that your organisation is prepared to adapt to these transformations. To learn how your organisation can navigate the implications of Colombia’s AI Policy and leverage opportunities in this evolving landscape, please contact Geusseppe Gonzalez at [email protected]
