11-7-2023 (MEXICO CITY) The Costa Rican Congress has turned to artificial intelligence to address the need for regulating AI technology. Lawmakers asked ChatGPT, an AI chatbot, to draft a law that aligns with the constitution and reflects legal expertise. The resulting text was sent verbatim to the legislature, sparking both positive reactions and concerns.
Congresswoman Vanessa Castro, leading the bill’s introduction, acknowledged the mixed response, stating, “We have had many positive reactions and many people who thought it to be very risky.”
ChatGPT recommended the creation of an institution in Costa Rica responsible for regulating AI systems. The proposed regulations would be based on principles such as accountability, explainability, bias prevention, and protection of human rights.
Although the bill was introduced in May, it is currently undergoing public discussions before it proceeds to the parliamentary commission for amendments and further debate in Congress.
Castro remarked that the experience taught them that AI is a legislative tool that still requires human intervention. “We learned that artificial intelligence is just another legislative tool that still needs the human hand,” she said.
Costa Rica joins the ranks of eight Latin American countries that have either discussed or approved laws to regulate AI in the past year.
Costa Rican Congresswoman Johana Obando expressed support for AI regulation but opposed the bill, claiming that ChatGPT fabricated statistics and articles from the Costa Rican constitution. She also criticized the bill for being a mere “list of good wishes” lacking substantive measures, a common concern in AI legislation debates across Latin America.
Obando stated, “ChatGPT said ‘we should regulate based on fundamental rights and international conventions.’ But what are those rights and conventions? The bill does not mention them.”
Latin American lawmakers draw inspiration from the European Union’s AI Act, which includes provisions prohibiting the use of AI technology in biometric surveillance and emphasizing transparency in AI-generated content.
In Mexico, a bill introduced in March aims to establish an ethical framework for AI development based on the protection of human rights and personal data. However, the draft lacks a description of the framework itself, and it is currently being discussed in public forums with experts and lawmakers.
Meanwhile, the Peruvian Congress has already approved the first AI regulation law in the region, awaiting the president’s signature for it to take effect. The law designates a national authority responsible for overseeing AI development, guided by principles of digital security and ethics.
Peruvian Congressman José Cueto, who spearheaded the bill, emphasized that the legislation is just a small part of a larger national strategy required for cybersecurity and data protection. “The heart of the law is to create an environment in which we can make an ethical, transparent, and sustainable use of AI,” Cueto, a cybersecurity expert and former admiral, stated.
In Brazil, AI regulation has been a topic of intense debate over the past four years, with three bills pending in Congress. One of the approved bills focused on principles but lacked enforcement mechanisms, leading to the creation of a Senate commission. This commission published a 900-page report proposing risk-based regulation, recommending the prohibition of AI systems that may harm individuals or target marginalized populations.
Concerns have been raised by anti-racism advocates, like researcher Tarzício Silva, regarding the exclusion of minority perspectives from the debate. Silva expressed worries about the use of facial recognition systems leading to disproportionate arrests among marginalized populations and automated hiring systems perpetuating discrimination against racial minorities.
The regional discussions on AI regulation emphasize the importance of combating bias and discrimination in AI systems. However, proposed legislation often lacks clarity on preventive measures, investigations, and penalties for such issues.
A common theme throughout the region is the aspiration to create an environment that fosters local experimentation and competition with multinational corporations like Microsoft and Google. In Brazil, one of the proposed bills suggests governmental authorization for the creation of a “regulatory sandbox,” providing a controlled framework for local businesses to experiment with AI technology.
Francisco Garijo, president of the Ibero-American Society of Artificial Intelligence, stressed the need for promoting the development of local AI products to counteract the dominance of American multinationals. Garijo stated, “We are currently colonized by products of a few American multinationals. The best way to deal with this colonialism is to promote the development of local products that can compete with them.”
Furthermore, experts believe that AI systems must be created specifically for Latin Americans, considering their languages, cultures, and active involvement in research and development.
At a regional AI summit held in Uruguay’s capital in March, experts who endorsed the Montevideo Declaration on Artificial Intelligence emphasized the importance of valuing Latin American participation beyond data collection and manual annotation.
“In the creation of AI technologies for the region, it is necessary to value their (Latin American) participation in research and development, and not only as mere producers of raw data and manual annotation,” the experts stated.
