There was uproar when politicians in the Brazilian city of Porto Alegre discovered that a newly enacted law was secretly written by OpenAI’s popular chatbot, ChatGPT. The experimental ordinance, aimed at preventing taxpayers from being charged to replace stolen water consumption meters, was revealed to have been crafted entirely by AI. City councilman Ramiro Rosário had asked ChatGPT to create the proposal, presented it to his peers without disclosing its origin, and the council unanimously approved it. The revelation has sparked objections and raised concerns about the role of AI in public policy. While some see the use of AI-powered chatbots as a promising tool, others worry about the unintended consequences of relying on machines to handle tasks currently performed by humans.
The law was enacted in October in Porto Alegre, making it potentially the first legislation in Brazil to be entirely written by artificial intelligence. Councilman Rosário made use of ChatGPT to draft the proposal, without making any changes or notifying his fellow council members about its origin. If he had revealed its AI authorship beforehand, Rosário believes the proposal wouldn’t have even been brought to a vote.
In response to the revelation, council president Hamilton Sossmeier initially expressed concern, calling it a dangerous precedent. ChatGPT and similar chatbots function by predicting the next word in a sentence, and they can sometimes generate false or misleading information. While they can be useful in simplifying tasks, they may not possess the same level of understanding and judgment as human lawmakers, particularly when it comes to interpreting complex legal principles and precedents.
The use of AI in lawmaking has seen limited experimentation worldwide. In Massachusetts, Democratic state Senator Barry Finegold enlisted ChatGPT to help draft a bill aimed at regulating AI models, including ChatGPT itself. However, Finegold emphasizes the importance of transparency, ensuring that everyone is aware when AI tools are utilized in the process.
In the case of Porto Alegre, there was no such transparency. Council President Sossmeier learned about the AI involvement when Councilman Rosário publicly boasted about it on social media. Initially opposed to the method, Sossmeier changed his stance after further research, acknowledging the inevitability of using AI in drafting legislation.
The revelation has ignited a debate about the use of AI in lawmaking and the extent to which it can effectively account for the complexities of legal frameworks. While chatbots like ChatGPT offer efficiency and speed in searching and citing existing laws, there is a consensus that more rigorous legal analysis often requires human expertise.
Councilman Rosário intentionally kept the AI origin of the proposal secret, aiming not only to address a local issue but also to spark a broader discussion on the potential of AI for good. He firmly believes that technology, when wielded responsibly, can lead to positive advancements.
As the world grapples with the implications of AI-powered chatbots, it becomes crucial to strike a balance between leveraging their capabilities and ensuring human oversight to safeguard against unintended consequences. The case of Porto Alegre serves as a reminder of the ethical and practical considerations that arise when AI becomes involved in shaping public policy.
Porto Alegre’s experiment with AI-written legislation showcases both the potential and limitations of using chatbots like ChatGPT in lawmaking processes. As technologies continue to evolve, the global conversation around their use in governance intensifies – a conversation that will shape the future of AI’s role in public policy and legislation.
