A municipality in Brazil passes a decree that was covertly authored by ChatGPT.

City legislators in Brazil have passed a law that seems to be the country’s first ever written by artificial intelligence, even though they were not aware of it at the time.

In October, the city council of Porto Alegre in the southern Brazil passed an experimental law. This week, councilman Ramiro Rosário disclosed that the law was drafted by a chatbot, leading to objections and raising concerns about the use of artificial intelligence in shaping public policy.

Rosário told The Associated Press that he asked OpenAI’s chatbot ChatGPT to craft a proposal to prevent the city from charging taxpayers to replace water consumption meters if they are stolen. He then presented it to his 35 peers on the council without making a single change or even letting them know about its unprecedented origin.

Rosário stated to the AP over the phone on Thursday that if the information had been disclosed earlier, the proposal would not have been put to a vote. The council, consisting of 36 members, unanimously approved it and the ordinance was implemented on November 23rd.

He stated that it would be unjust to the public to take the chance of the project not receiving approval solely because it was created by artificial intelligence.

The emergence of ChatGPT in the market only a year ago has stirred up a worldwide discussion about the potential effects of groundbreaking AI-driven chatbots. While some view it as a hopeful resource, it has also raised worries and unease about the unintentional or unwanted consequences of machines taking over tasks currently done by humans.

Porto Alegre, with a population of 1.3 million, is the second-largest city in Brazil’s south. The city’s council president, Hamilton Sossmeier, found out that Rosário had enlisted ChatGPT to write the proposal when the councilman bragged about the achievement on social media on Wednesday. Sossmeier initially told local media he thought it was a “dangerous precedent.”

AI chatbots like ChatGPT use large language models to generate responses by continually attempting to predict the next word in a sentence. However, these models are susceptible to creating incorrect information, which is known as hallucination.

According to a recent study by Vectara, chatbots may occasionally include incorrect information when summarizing a document. The rate of this occurrence varies, with the most advanced GPT model having a false information rate of about 3%, while one of Google’s models has a rate of around 27%.

Andrew Perlman, dean of Suffolk University Law School, wrote in an article on the website of Harvard Law School’s Center of Legal Profession that the introduction of ChatGPT could signal a significant change, possibly even more significant than the rise of the internet. However, he also cautioned about potential drawbacks.

According to Perlman, ChatGPT, being a machine learning system, may not be capable of fully comprehending the intricacies and intricacies of the law. Due to this limitation, it may not possess the same level of comprehension and judgement as a human lawyer when it comes to interpreting legal principles and previous cases. This could potentially cause issues in situations where a more comprehensive legal analysis is necessary.

Other individuals have also utilized ChatGPT to a lesser extent or with less favorable results, preceding Rosário of Porto Alegre as the initial lawmaker to test its capabilities.

Democratic state Senator Barry Finegold from Massachusetts enlisted the assistance of ChatGPT in drafting a bill that would oversee the use of artificial intelligence models, including ChatGPT. The bill was submitted earlier this year but has not yet been voted on.

According to Finegold, ChatGPT can be utilized to assist with the arduous tasks involved in creating laws. This includes efficiently and accurately searching for and referencing existing laws. He emphasized the importance of transparency by ensuring that all parties are aware of the use of ChatGPT or a comparable tool in the process.

“We advocate for the watermarking of ChatGPT-generated work,” he stated, acknowledging that the implementation of artificial intelligence in creating legislation is unavoidable. “I support the use of ChatGPT for drafting bills, as long as it is clearly indicated.”

Rosário failed to provide transparency in her proposal in Porto Alegre, as Sossmeier revealed that she did not disclose ChatGPT’s involvement in writing the proposal to other council members.

The decision to keep the source of the proposal undisclosed was deliberate. Rosário informed the AP that his goal was not only to address a specific issue in his community, but also to initiate a discussion. He explained that he inputted a 49-word prompt into ChatGPT and it quickly generated the complete proposal, including justifications.

He stated, “I strongly believe that there will be a new technological revolution for humanity.” He acknowledged that the tools we have created as a society can be utilized for both positive and negative purposes. Therefore, it is crucial to demonstrate their potential for good.

The leader of the council, who originally criticized the approach, now seems to have changed their mind.

Sossmeier stated that they have reconsidered their previous statement. After delving deeper into the subject, they came to the realization that this will likely become a popular trend, whether it is unfortunate or fortunate.


Savarese wrote this report from Sao Paulo. The Associated Press reporters, Steve LeBlanc in Boston and Matt O’Brien in Providence, Rhode Island, also contributed to this article.

Source: wral.com