Title: Latin America Works to Regulate AI as ChatGPT Drafts New Laws
Latin American countries are engaging in a heated debate over the regulation of artificial intelligence (AI), with the aim of preventing potential harm and promoting local innovation. Critics, however, argue that Indigenous and Black perspectives have been excluded from these discussions. In an effort to address this, lawmakers in Costa Rica recently turned to ChatGPT, an AI chatbot, to draft a new AI regulation bill.
Lawmakers in Costa Rica instructed ChatGPT to think like a lawyer and draft a bill that adhered to the principles outlined in the country’s constitution. The resulting text was then presented to the legislature. While some expressed concerns about relying on an AI chatbot for such an important task, Congresswoman Vanessa Castro, who spearheaded the bill introduction, noted that they received mostly positive reactions.
ChatGPT’s recommendations for Costa Rica included the creation of an institution to regulate AI systems, with a focus on accountability, explainability, bias prevention, and the protection of human rights. Although the bill was introduced in May, it is currently being discussed in public forums before undergoing further debate and amendments in Congress.
This move by Costa Rica makes it the eighth country in Latin America to initiate discussions or approve laws to regulate AI in the past year. However, not all lawmakers are in agreement. Congresswoman Johana Obando supported the need for AI regulation but opposed the bill that ChatGPT drafted, arguing that it included made-up statistics and failed to reference specific rights and conventions.
Latin American lawmakers are drawing inspiration from the European Union’s AI Act, which includes provisions banning the use of AI in biometric surveillance and emphasizes the importance of transparency regarding AI-generated content. Mexico, for instance, introduced a bill in March that encourages the development of an ethical framework for AI, based on the protection of human rights and personal data.
In June, Peru became the first country in the region to pass an AI regulation law, pending the president’s signature. The law designates a national authority responsible for overseeing AI development, emphasizing digital security and ethics.
AI regulation in Brazil, on the other hand, has faced ongoing debates over the past four years, with multiple bills awaiting congressional approval. One bill, approved by the House of Representatives in 2021 but blocked by the Senate, focused mainly on principles and lacked enforcement mechanisms. This led to the creation of a Senate commission that published a comprehensive report proposing risk-based regulations, banning AI systems that could harm marginalized populations.
However, critics argue that perspectives from racial minorities, including Black and Indigenous communities, have been largely excluded from these discussions. To address this, a new bill based on the existing proposals is currently being discussed in the Senate, with a parliamentary commission responsible for its further examination.
Key concerns in the region revolve around preventing the use of facial recognition systems that could lead to the disproportionate targeting and arrest of marginalized populations, as well as addressing biases in automated hiring systems. Advocates also emphasize the importance of people’s right to review algorithmic decisions and the potential for reparations for those harmed by AI systems.
Another critical aspect of the AI regulation discussion is the promotion of local experimentation and innovation. Many lawmakers argue that Latin America needs to create an environment that fosters local competition with multinational corporations like Microsoft and Google. Suggestions include the establishment of a regulatory sandbox that allows local businesses to experiment with AI technologies in a controlled setting.
Experts from the Ibero-American Society of Artificial Intelligence stress the need for AI systems to be developed specifically for Latin America, reflecting the languages, cultures, and unique challenges in the region. They believe that this will help address the dominance of foreign multinational corporations and promote the development of local AI products.
As Latin America continues its journey towards AI regulation, lawmakers are grappling with the task of crafting effective and inclusive laws that protect citizens’ rights and prevent potential harm. The diverse perspectives brought to these discussions will be crucial in shaping a comprehensive regulatory framework that promotes innovation while safeguarding the region’s values and interests.
