How AI Could Make Our Laws: First Bill Written by ChatGPT Gets Approved

KEY TAKEAWAYS

A bill written by ChatGPT was approved in Brazil for the first time in the world. Is AI writing laws simply another application of modern AI, or a change that could revolutionize our world? Would you vote for ChatGPT?

In Porto Alegre, Brazil, the first legislation written by Artificial Intelligence (AI) has passed.

ChatGPT prepared the bill in just a couple of minutes, and all 36 city council members unanimously voted for it — without even knowing an AI wrote it. Although a small city law may seem a small thing, this is a first step into a new era.

An era where an AI can churn out legislation in 15 seconds, and 36 committee members can say: “Fine by us.”

The implications of this small step may be world-changing. Might politicians be the ones who become obsolete and lose their jobs to AI?

That’s not likely to happen, even if some quarters may wish for it! But we do have some caveats to point out if chatbots will help rule the world.

The Law That Took 15 Seconds to Write

October 2023. With its 1.3 million population, Porto Alegre is the second-largest city in southern Brazil. Councilman Ramiro Rosário was preparing to write down a new legislation that would exempt the city’s residents from having to pay stolen water meters from their own pockets.

Advertisements

His constituents didn’t want to be charged by the Municipal Water and Sewage Department whenever their water meters were stolen.

Rosário knew all too well that sitting down with the other 35 members of the commission and the legal sources would have taken days in the best-case scenario. So he opened up the large language model (LLM) and wrote down a simple prompt:

Create a municipal law for the city of Porto Alegre, originating from the legislature and not the executive, which prohibits the Municipal Water and Sewage Department from charging the property owner for the payment of a new water meter when it is stolen.”

The chatbot took just 15 seconds to write down an eight-part bill that was perfectly sound, and it was so bright as to set some additional rules.

For instance, the city had 30 days to substitute the stolen meters, and when it couldn’t meet that deadline, property owners were exempted from having to pay their water bills.

The legislation proposal was put in front of the other 35 council members, who promptly approved it unanimously. So far so good, right? Well, not so much actually, since Rosário’s peers had no idea a chatbot wrote the law.

After Rosário exposed the truth on his social profile, reactions were mixed. Some were fascinated by the idea; others were outraged by the lack of transparency; others claimed that the law set a dangerous precedent.

According to Rosário, this is nothing but an unavoidable step into the modernization of our world: “[This] is really no different than the changes the internet and computers brought. Remember how everyone was scared of computers taking over? Well, look at how helpful they’ve been so far.”

So, are there no downsides to his revolutionary choice?

The Risks of a World Where Chatbots Write Laws

Rosário’s law is the first one approved, but it’s not the first one that was written by an AI.

Earlier this year, Senator Barry Finegold from Massachusetts employed ChatGPT’s help to prepare a bill regulating AI models, including ChatGPT itself.

READ MORE:

The bill has yet to be voted on, but unlike its Brazilian counterpart, everybody is perfectly aware that the chatbot wrote it. All things given, it is almost sure that employing the help of generative AI will become a trend, eventually, mainly because it speeds up what is otherwise a long and tedious process.

However, this trend requires its own laws to be regulated.

While it is easy to believe Rosário’s words when he says he kept this a secret to spark a debate, we must expect transparency about AI use to be required in lawmaking.

In fact, Sen. Finegold wanted this type of content to be watermarked, which may be necessary, as chatbots are not perfect.

Chatbots often spout (mis)information that ranges from “somewhat wrong” to “completely false,” and it comes with the territory that something like this can’t be accepted when it comes to legislation.

While later models could undoubtedly reduce this risk in the future, human oversight will still be needed, so it’s not like lawyers and politicians will lose their place anytime soon.

So, put the champagne down because a world where AI would write all laws won’t be a world without politicians.

AI and Human Oversight

Inaccuracies, falsehoods, biases, and all these mistakes need a solution to fix them, but who will decide what the correct answer is?

By far, the most significant risk of a world where AI writes laws on its own is that we know nothing of how AI writes these laws. The “black box” nature of most generative AI algorithms means that, right now, no governmental agency can make a risk assessment and take the appropriate security measures.

Unless Explainable AI becomes more established, which can shed light on the decision process on the route to an answer, the “black box” and inability to “show your workings” may not sit easily in the realm of lawmaking.

We live in a world where if information cannot be found on a search engine, it’s like it never existed in the first place.

One last food for thought: If chatbots are regularly used to write laws, tech companies would have access to governmental data, which could directly or indirectly affect the lawmaking process.

This is an incredible amount of power that could be driven into their hands without any effort on their side, and it must be regulated before it is too late. And humans must hold it while we still have time to do that.

The Bottom Line

AI writing laws are, quite simply, another application of modern AI in generating advanced texts.

What’s not routine, however, is the implications of having a few Big Tech AI controlling (or even monopolizing) the lawmaking process.

It’s early, but the world moves quickly, and we are delving into dangerous territory — we must have a conversation with ourselves and then a framework to ensure AI and lawmaking are transparent, balanced, and unbiased to the best of human possibilities.

Advertisements

Related Reading

Related Terms

Advertisements
Claudio Buttice

Dr. Claudio Butticè, Pharm.D., is a former Pharmacy Director who worked for several large public hospitals in Southern Italy, as well as for the humanitarian NGO Emergency. He is now an accomplished book author who has written on topics such as medicine, technology, world poverty, human rights, and science for publishers such as SAGE Publishing, Bloomsbury Publishing, and Mission Bell Media. His latest books are "Universal Health Care" (2019) and "What You Need to Know about Headaches" (2022).A data analyst and freelance journalist as well, many of his articles have been published in magazines such as Cracked, The Elephant, Digital…