Artificial Intelligence and New Technologies

Yes. There is currently no specific artificial intelligence law in force, but that does not mean its use is free of rules. Already applicable are, among others, Law No. 21,719 (Ley N° 21.719) on personal data protection (relevant when AI processes information about individuals), the consumer protection rules of Law No. 19,496 (Ley N° 19.496), and the intellectual property regime of Law No. 17,336 (Ley N° 17.336). In addition, a bill is under discussion in Congress. The sensible course is therefore not to wait for the new law, but to use AI under a clear internal policy that defines what data may be uploaded, what the AI is used for, and who supervises the results.

Many companies already use generative AI tools without clear rules, which creates risks: leaks of confidential information or personal data, erroneous outputs used in decision-making, and intellectual property problems. An internal AI use policy defines which tools are permitted, what information may or may not be uploaded, how outputs are reviewed before being used, and who is responsible for each use. This helps comply with Law No. 21,719 (Ley N° 21.719) on data protection and anticipate the requirements of the future AI law. It is a simple governance measure that prevents bigger problems and brings order to the adoption of these technologies.

A bill regulating artificial intelligence systems is currently before Congress (Bulletin 16.821-19). Its approach is risk-based: it distinguishes systems of unacceptable risk (prohibited), high risk (with reinforced obligations such as risk management, human oversight, and documentation), limited risk (with transparency duties), and no evident risk. It also addresses the transparency of AI-generated content and provides for penalties. Because it is still a bill, its content may change during the legislative process, but it is worth following closely and preparing the organization for its requirements.

Uploading personal data of clients, employees, or third parties to an AI tool is a form of data processing and is subject to Law No. 21,719 (Ley N° 21.719). This requires having a lawful basis, informing the data subjects, limiting use to the declared purpose, and making sure the tool’s provider offers adequate safeguards, especially if the information is processed abroad. Uploading sensitive or confidential data to services without a contract or controls can create liability and leaks that are difficult to reverse. Before adopting a tool, it is advisable to review its terms, where the data is stored, and what the provider does with it.

This is an open and evolving question. Law No. 17,336 (Ley N° 17.336) on intellectual property protects works of human creation, so the authorship of content generated entirely automatically by an AI can be debatable. Added to this are the terms of use of each tool, which govern who owns the outputs and under what limitations they may be used commercially, and the risk that the content may reproduce third-party works. In practice, it is advisable to review the provider’s terms, keep a traceable record of human involvement in the creation, and avoid using outputs that could infringe the rights of others.

With caution. Law No. 21,719 (Ley N° 21.719) grants individuals rights with respect to automated decisions that significantly affect them, including profiling, which may entail the right to human intervention and to know the logic involved. Using AI for decisions on credit, hiring, dismissal, or customer service therefore requires transparency, genuine human oversight, and the ability to review and correct the outcomes. The AI bill under discussion aims to reinforce these requirements for higher-risk uses. Before automating a sensitive decision, it is advisable to assess its impact and always keep effective human control in place.