- ChatGPT was used to draft legislation regulating AI in Costa Rica.
- Many countries in Latin America are debating laws regulating AI.
- Promoting regional innovation can prevent “AI colonialism.”
- For more articles, see Tech and Trends. home page.
When Costa Rican lawmakers wanted to draw attention to the need to regulate artificial intelligence, they asked ChatGPT to create a new law.
Lawmakers told the chatbot to “think like a lawyer” and draft legislation in accordance with the Constitution. They then sent the resulting document verbatim to Congress.
Rep. Vanessa Castro, who led the bill’s introduction, told the Thomson Reuters Foundation that “there was a lot of positive response and many people thought it was very dangerous.”
ChatGPT recommended that Costa Rica create a body to regulate AI systems based on the principles of accountability, accountability, anti-bias and human rights protection.
The bill, which was introduced in May, is currently under public debate before being submitted to parliamentary committees for revision and further debate in parliament.
“We’ve learned that artificial intelligence is just one legislative tool that still needs a human hand,” Castro said.
Costa Rica is the eighth country in Latin America to discuss or approve legislation regulating AI in the past year.
Costa Rican Congressman Johanna Ovando said she supported AI regulation but opposed the bill, saying ChatGPT only fabricated statistics and articles on the Costa Rican constitution.
But her main objection is to say that the bill is just a “list of goodwill” that doesn’t mesh well, a common concern with AI bills being debated across Latin America.
ChatGPT “should be regulated based on fundamental rights and international conventions,” Obando said. “But what are those rights and customs? The bill doesn’t mention them.”
Latin American lawmakers are pushing for regulation motivated by the European Union’s AI law, which includes rules banning the use of technology in biometric surveillance and clarifying AI-generated content. is
In Mexico, a bill submitted in March encourages the creation of an ethical framework for AI development based on human rights and the protection of personal data.
However, the ethical framework itself was not included in the draft and is currently being debated publicly by experts and lawmakers.
Peru’s parliament approved the region’s first law regulating AI in June, but it’s just waiting for the president’s signature to take effect.
The law designates a national agency to oversee the development of AI on the principles of digital security and ethics.
Peruvian Congressman Jose Cueto, who led the bill’s introduction, said the bill was just one piece of a much-needed national strategy on cybersecurity and data protection.
Cueto, a cybersecurity expert and former Admiral of the Navy, said, “At the core of this law is creating an environment where AI can be used ethically, transparently and sustainably.”
racism and discrimination
Brazil has had a heated debate over AI regulation for the past four years, with three bills pending in Congress.
Targisio Silva, an algorithm bias researcher and Fellow of the US nonprofit Mozilla Foundation, said one of the AI legislative frameworks approved by the House but blocked by the Senate in 2021 focuses primarily on principle. and lacked an enforcement mechanism. organization.
The bill led to the creation of a Senate committee, which released a 900-page report proposing risk-based regulations, including those that would harm people or target marginalized groups. AI systems that may
For example, the report recommends banning the use of AI systems for social scoring, which conditions access to public services based on scores obtained for people’s behavior.
But anti-racists like Silva worry that the debate excludes minority perspectives.
“The commission consisted of 18 jurists and 80 experts, not a single Brazilian racial minority,” Silva said. “They didn’t care about blacks and indigenous peoples.”
AI regulation has become a hot topic in Brazil, and a new bill based on three bills pending in Congress was introduced to the Senate in May and will soon be considered by a parliamentary committee.
For Silva, a key element of the debate is blocking the use of facial recognition systems, which enable the wrongful arrest of marginalized people, and automated employment systems that discriminate against racial minorities.
Silva said of the possibility of compensation for those harmed by AI systems, saying, “We advocate the right to review the decisions of the algorithms, and we are discussing what will be the entitlement to reparations.” Stated.
Legislators in the region agree that fighting stigma and discrimination in AI systems should be at the heart of new regulation, but much of the proposed legislation aims to prevent it and investigate it. and are vague about how to punish them.
AI colonialism
A common element in the regional debate on AI regulation is the need to create an enabling environment that fosters local experimentation and allows competition from multinationals such as Microsoft and Google.
In Brazil, one of the bills proposes government authorization to create a “regulatory sandbox,” a framework that allows local companies to experiment with AI technology in a controlled environment.
“We are now a colony of the products of several American multinationals,” says Francisco Garrijo, president of the Ibero-American Association for Artificial Intelligence, which brings together experts from across the region.
“The best way to deal with this colonialism is to promote the development of local products that can compete with colonialism,” he said.
Experts who gathered at a regional AI summit in the Uruguayan capital in March said AI systems would be tailor-made by and for Latin Americans, taking into account their language and culture. He also said he needed it.
Experts who signed the Montevideo Declaration said, “In creating AI technologies for the region, we need to value[Latin America’s]participation in research and development, not just as producers of raw data and manual annotations.” there is,” he said. artificial intelligence.
