loader image

Foto: Alexandra Koch de Pixabay

Global and regional developments in the regulation of Artificial Intelligence

14 December, 2023 | Ricardo Changala

Ramiro Rosario, a counselor in the Municipal Chamber of Porto Alegre, State of Rio Grande do Sul, Brazil, aimed to propose a rule exempting taxpayers from paying compensation for the theft of hydrometers. Perhaps out of technological curiosity, time-saving considerations, annoyance, or a combination of these reasons, he asked the ChatGPT tool to prepare a draft municipal law. According to Rosario, the virtual assistant drafted a text with eight articles, including its justification and suggestions for changes to the initial proposal made by the councilor. Presenting the initiative as received from ChatGPT, the councilman gained favorable opinions from local legislative authorities, and finally, on October 18, 2023, it was approved by the plenary of the Municipal Chamber. Rosario shared how the project unfolded only after the legislative process ended, on November 23, when the mayor of Porto Alegre, Sebastiao Melo, adopted the new law.

Once the facts became known, a debate arose in the municipality of Gaúcho. While Rosario defends the actions taken, stating that artificial intelligence will enhance productivity, the president of the Municipal Chamber pointed out that this occurrence sets a dangerous precedent. This could lead to unforeseen impacts, especially in more complex laws, emphasizing the necessity for a legal framework governing artificial intelligence.

Other opinions have focused on diminishing the significance of the way a bill is drafted, highlighting that human presence is crucial during debates and law approvals in the legislative body.

Whether acknowledged or not, similar situations to the one described are likely occurring in numerous public institutions and private entities worldwide. This strongly indicates the need to establish suitable legal frameworks to regulate the increasingly advancing field of artificial intelligence.

This realization has been acknowledged for years by international organizations and several States, which have made strides in addressing this issue. At the global level, the AI principles adopted by the Artificial Intelligence Commission of the Organization for Economic Cooperation and Development (OECD) and UNESCO’s Recommendation on the Ethics of Artificial Intelligence in 2021 stand out. These provide a universal framework of values, principles, and actions guiding States and private entities in formulating their AI-related laws, policies, or instruments, aligning with international law to ensure ethical incorporation across all AI system life cycles.

The European Union has made significant strides in this regard. On December 9, 2023, the Presidency of the European Council and negotiators of the European Parliament provisionally agreed upon a proposal establishing harmonized AI standards, termed the “artificial intelligence law.” This initiative aims to ensure the safety of AI systems in the EU, respecting fundamental rights and EU values, while encouraging investment and innovation in AI within Europe.

The core idea is to regulate AI based on its potential societal harm using a risk-based approach: stricter rules for higher-risk AI. The initiative covers regulation on high-impact AI models, a governance system with enforcement powers at the EU level, and requirements for accessing the EU market.

These requirements have been clarified and adjusted to be technically feasible and less burdensome for stakeholders to comply with, such as in terms of data quality or the technical documentation needed by small and medium-sized enterprises to demonstrate compliance with high-risk AI system requirements.

The provisional agreement prohibits various actions, including cognitive-behavioral manipulation, non-selective collection of facial images from the internet or CCTV recordings, emotion recognition in workplaces and educational institutions, social scoring, biometric categorization to infer sensitive data, and certain instances of predictive policing for individuals.

It also forbids scoring based on social behaviors or personal characteristics, AI systems manipulating human behavior to bypass free will, and the exploitation of people’s vulnerabilities using AI, such as due to age, disability, or social and economic situations.

Regarding specific countries, a recent study highlights various government efforts toward AI regulation and its integration into public service technology. For instance, the Republic of Korea integrated AI into its Government Digital Platform, and the United Kingdom strives to enhance its National Health Service by supporting research and innovation in AI-based disease detection technologies.

The mentioned report presents a global ranking on AI development, revealing that Latin America and the Caribbean (LAC) hold an average score of 41.50, ranking sixth among nine covered regions. Significant disparities exist within the region: Brazil leads, nearly 42 points ahead of Haiti, which holds the last position with 21.97.

Five countries—Brazil, Chile, Uruguay, Colombia, and Argentina—stand out with scores ranging between 63.70 and 57.72 points. The Dominican Republic recently distinguished itself as the first Caribbean country to launch its National AI Strategy. Cuba, Jamaica, and Mexico also announce forthcoming concrete actions on this issue.

In 2023, a notable development in the region was the signing of the Santiago Declaration by 20 governments of Latin America and the Caribbean. This emerged as the main outcome of the Summit of Ministers and High Authorities of Latin America and the Caribbean on Artificial Intelligence, organized by CAF (Development Bank of Latin America), UNESCO, and the government of Chile.

The declaration recognizes the necessity for proactive engagement by governments to harness AI’s opportunities while addressing its risks. This marks a significant step in regional collaboration to prepare for AI.

In the Declaration, the States decided:

  • To deepen regional dialogue concerning AI development and deployment, reflecting the needs and interests of Latin America and the Caribbean.
  • To approve the establishment of a Working Group towards constituting an intergovernmental Artificial Intelligence Council for Latin America and the Caribbean. This aligns with the UNESCO Recommendation on the Ethics of AI, aiming to bolster regional capacities in this field.

Given the urgency and immense importance of the issue, the hope is that the agreements reached in Santiago will lead to effective actions, giving them substantial meaning and impact.