With the issuance of CONPES 4144 of 14 February 2025, it cannot be denied that the National Council for Economic and Social Policy has successfully carried out in-depth research, well supported by studies on the impact of artificial intelligence (AI) in different national spheres, and that on this basis it has formulated the National Artificial Intelligence Policy in accordance with the economic and social needs of Colombia, to promote its responsible use and harness its potential for the growth and transformation of the country.
With regard to the issue of intellectual property (IP) and its relationship with AI, it establishes a strategy that addresses important policies such as the transfer of technology from the academic to the productive sector, the strengthening of institutional capacities for the protection of intellectual property and mechanisms to mitigate ethical risks, which we feel are of the utmost importance.
One positive aspect of the policy, apart from its comprehensive vision of the complexity of IP in relation to AI, is that it wisely sets out the route to achieving the updating of the regulatory framework for intellectual property related to artificial intelligence by 2030, in line with technological advances and international best practices. Another positive aspect is the action of training the state entities involved to generate more specialised knowledge on artificial intelligence and intellectual property. It also proposes the development of digital skills and talent and seeks to strengthen institutional and human capacities in the field of AI. It is to be hoped that the strengthening of these capacities will be extended or will give a glimpse of the importance of IP, which is the basis of development and growth, but which continues to be stigmatised in an erroneous appreciation of it as a barrier to the dissemination and appropriation of knowledge.
However, there is a vacuum, as policy fails to achieve a greater degree of precision in Colombia’s position on the ownership of IP rights for creations generated with the support of artificial intelligence. Thus, we are referring to the lack of a position as to who would be the owner when AI creates content, or whether the country would adopt a model of co-authorship of human and machine, or perhaps that in some circumstances these results would fall into the public domain, following to some extent the pattern of those inventions that, although they could be protected by patents, for one reason or another almost always of collective interest, are not granted such patent rights.
The complexity of intellectual property in relation to the results of AI seems to be a recurring theme in publications and articles on Artificial Intelligence, but it is a reality and is recognised in the document (CONPES 4144). For this reason, it establishes strategies for exhaustive analysis of the risks associated with intellectual property in relation to AI, as well as a regulatory strategy to provide clear and structured legislation that meets the needs of intellectual property and the enforcement of IP rights, with legislation that adapts to international standards. And in the same way, a strategy for mitigating IP-related risks that guarantees that said IP will not be an obstacle to democratising knowledge or to SMEs’ access to technological innovations.
Looking more tangibly at where we are going, we find that the document does not mention whether Colombia will adopt intellectual property models from other countries that are already in the process of establishing AI and IP regulations, such as China, the United States and the European Union. Perhaps the CONPES document was not the right place to define it with such precision, but on the road to implementation it would have been useful to provide a strategic guideline for international cooperation that would allow for the exchange of good practices and thus facilitate the harmonisation of regulation in this area.
It would have been ideal if some guidelines had been provided following the developments or advances made by other countries which, although not definitive, serve as examples to be followed, such as the United States, a country that does not grant IP rights to works generated 100% by artificial intelligence because it considers human creativity to be an essential requirement for protection under copyright law and is in the process of defining how it will regulate human-AI co-authorship and the protection of AI-generated inventions.
It is worth mentioning here that, for its part, the European Union demands transparency in AI-generated content and regulates its training with protected works. The United Kingdom is considering granting copyright when there is human intervention and possible licences for the use of protected content. China allows AI-assisted patents, and Japan, with a more flexible approach, allows AI to be trained with protected content without authorisation, except for commercial purposes.
On the other hand, and paying attention to the organic component, we will have to refer to the role that the CONPES document rightly assigns to the Cross-sector Commission on Intellectual Property (CIPI) created in 2010. It is the coordinating body providing senior guidance to the National Intellectual Property Administrative System and plays an important role in the implementation of CONPES 4144 policy. However, it would have been advisable to designate a specific entity or person responsible for following up on each of the initiatives and guaranteeing compliance with the strategies. Bear in mind that according to the first paragraph of Article 5 of Decree 1162 of 2010, the Director General of the National Copyright Directorate, the Director of the ICA (Colombian Agricultural Institute), the Director of INVIMA (National Institute for Food and Drug Monitoring) and the Superintendent of Industry and Commerce are part of the cross-sector commission with the right to speak but not vote. As a result of these technological advances, it would be important for the preservation of the importance of intellectual property that these parties should have a more active legal participation in CIPI’s decisions. Of course, they are familiar with the day-to-day running of the sectors under their responsibility and can better identify needs and define solutions.
It is positive that the document establishes a continuous measurement and review of how artificial intelligence affects intellectual property that arises from it or with its intervention, or that affects that which is already granted, as this allows for regulatory adjustments based on real data and situations, and the support of the CIPI guarantees the cross-institutional approach that every system needs.
On the other hand, the provision of technological tools to identify possible IP infringements arising from the use of AI is an excellent policy provision, and of course there must be opportunities for cooperation between government institutions, academia, and the private sector to achieve this regulatory development and the development of IP and AI strategies. Each of these parties is more familiar with their own needs and circumstances, so that everyone can contribute, and joint participation is necessary.
Another element is the promotion of IP protection with awareness campaigns to raise awareness among creators, and among artificial intelligence companies and developers, of how to protect their innovations. There would also be incentives for technology transfer, seeking to ensure that AI research and development is transferred and used in industry. This encourages innovation in various sectors, and at the same time encourages the registration of patents that arise from it, not only to protect the creations but to make the marketing and licensing of these technologies more secure and viable.
For this process to be truly effective, it is essential to design and implement a strategy that incorporates the ethical and sustainable use of artificial intelligence, ensuring that its development and application respect the principles of transparency, fairness and responsibility, in line with regulatory frameworks and international best practices.
All of the above, duly implemented, gives us peace of mind that we are on the right track, as it ensures that the institutions in charge of IP will have a greater capacity for action in response to the challenges of artificial intelligence. We are therefore set to issue regulations to protect creations generated with artificial intelligence while respecting existing human IP rights and those arising from AI. On an even more positive note, incentives will be created to enable companies and researchers to protect their innovations. Successful implementation!
Written by: Claudia Caro, Head of Colombia Office
