AI bill moves forward in Chile

The draft "Law Regulating Artificial Intelligence Systems", which was introduced in May 2024, continues to advance in the Chamber of Deputies.

The rapid progress that AI is experiencing makes it necessary for the legal system to address the risks and contribute to promoting the development and use of this technology, while at the same time safeguarding the fundamental rights of individuals.

Chile has been working since 2019 on a roadmap for the development of AI, redefining the National AI Policy and creating the National Centre for Artificial Intelligence (CENIA). Internationally, there have been several developments in this area. The European Union has been the first to pass an AI-specific law, which focuses on different levels of risk (unacceptable, high, limited, minimal), with unacceptable ones such as social scoring systems and manipulative AI being banned. In the US, self-regulation has prevailed so far, with voluntary commitments by technology companies to mitigate risks. And in China there are strict regulations on algorithms, synthetic content and generative AI.

Main outlines of the draft law

1. Risks

The text adopts a regulatory approach based on risk levels, similar to the model implemented in Europe's AI law. Four categories are distinguished:

• Unacceptable risk: systems prohibited for infringing fundamental rights.
• High risk: permitted under strict control rules.
• Limited risk: subject to transparency obligations.
• No obvious risk: no specific measures.

This classification allows regulation to be tailored to the severity and likelihood of the effects that AI systems may generate.

2. Protection of fundamental rights

AI uses are identified that could affect health, privacy, equality before the law and other essential rights. These include subliminal manipulation systems, biometric categorisation with sensitive data, generic social rating and emotional recognition in sensitive contexts.

Scope of application

The rule would apply to:

• Providers introducing AI systems on the market or putting them into service in Chile.
• Implementers of AI systems domiciled in Chile.
• Foreign suppliers and implementers, when the output information of the AI system is used in Chilean territory.
• Importers, distributors and authorised representatives of AI systems, domiciled in Chile.

4. Promotion of innovation

The project contemplates the creation of controlled test areas to encourage safe experimentation with AI systems before their commercialisation. In addition, specific support measures are promoted for smaller companies, facilitating their access to these spaces and their participation in governance bodies.

5. Guiding principles

The following general principles should be observed by all operators:

• Human intervention and supervision
• Transparency and explainability
• Robustness and technical security
• Privacy and data governance
• Diversity, non-discrimination and fairness
• Social and environmental welfare
• Protection of consumer rights

These principles guide both system design and public policy.

6. Institutionality and governance

The creation of the AI Technical Advisory Council is proposed, composed of representatives of the State, academia, industry and civil society. Likewise, the future Personal Data Protection Agency is assigned the function of supervising and sanctioning compliance with the law.

Status of the project

The bill is currently in the first constitutional procedure before the Future, Science, Technology, Knowledge and Innovation Committee of the Chamber of Deputies. It is being processed with simple urgency, which means that it must be discussed within a reasonable period of time, but not immediately.

Conclusion

Society as a whole - and Chile in particular - needs a legal framework that not only encourages the responsible development of AI, but also provides legal certainty to those who design, implement and use these technologies. At the same time, it is essential to protect people from the risks that may arise from improper or non-transparent uses of AI.

In this sense, the project seeks to balance innovation and protection by promoting flexible, risk-based regulation that allows for adaptation to technological change without losing sight of democratic values and fundamental rights.

Javier Sabido Intellectual Property Manager Ecija Chile.