In what seems like the plot to every cautionary science fiction ever, the United Arab Emirates has a bold new plan: let artificial intelligence write laws.
The UAE’s top officials say they want to turn over much of the legislative process — writing, updating, and reviewing laws — to AI. The idea is that this will lead to smarter rules, faster government and fewer bureaucratic roadblocks. It’s a world-first experiment in letting machines shape the legal code of an entire country.
“This new legislative system, powered by artificial intelligence, will change how we create laws, making the process faster and more precise,” said Sheikh Mohammed bin Rashid al-Maktoum, also the ruler of Dubai, for state media.
But while the pitch sounds slick, the implications are anything but simple.
Why the UAE is doing this
Legislative systems are notoriously slow, inefficient, and often just plain confusing. Drafting new laws can take years, and political deadlock can stall important reforms. The process also tends to make laws wordy and complex, making it sometimes extremely difficult for regular people to understand them.
In theory, AI can cut through that. It can scan thousands of laws in seconds, spot inconsistencies, identify gaps, and even borrow ideas from other countries with similar legal systems. That’s a powerful tool, no doubt. It can help lawmakers focus on actual problems rather than wasting time on paperwork or political theater.
This new system will be overseen by a government unit called the Regulatory Intelligence Office. The AI will pull from a huge database of local and federal laws, court decisions, and government records. It’ll analyze all that data and start suggesting legal fixes, tweaks, and updates. And yes, it will actually write the law in plain language, in multiple languages, so everyone in the country — from judges to expats — can understand it. In a country like the UAE, where 90% of the population is non-citizen, that could be transformative.
The country says the system could cut legislative drafting time by 70%. But critics warn the experiment could sacrifice transparency, fairness, and democratic oversight in the process.
This could go horribly wrong in more than one way
Here’s the catch: laws are about more than efficiency. They are about fairness, rights, and values. And that’s where things get dicey.
First of all, AI systems are still unreliable. Vincent Straub, an AI researcher at Oxford, told the FT that these systems still “hallucinate” facts and “have reliability issues and robustness issues.” In other words, they make stuff up.
Even worse, AI doesn’t understand human consequences. It doesn’t grasp context. And it definitely doesn’t understand justice. As Marina De Vos, a computer scientist at the University of Bath, put it, the AI could propose something that makes sense to a machine but may make absolutely no sense to really implement. It could end up deciding people’s lives on things that just don’t make sense in our society.
Then there’s the lobbying and abuse problem.
AI doesn’t just write perfectly balanced laws — it has its own biases and can even be weaponized to subtly manipulate them. In the U.S., researchers like Bruce Schneier and Nathan Sanders have warned that AI could create what they call “microlegislation”. These are small, hidden tweaks to bills that serve powerful interests without anyone noticing. Want to tilt the rules in favor of one company or industry? Feed your AI the right data, and it’ll quietly rewrite the playing field.
Everyone hates red tape. But red tape exists for a reason
There’s another reason why the UAE is trying this out: because it’s not a democracy. It’s an authoritarian state that has been described as a “tribal autocracy” where the seven constituent monarchies are led by tribal rulers in an autocratic fashion. There are no democratically elected institutions, and there is no formal commitment to free speech.
In other words, the UAE rulers don’t need to win votes or debate bills on the floor. That gives it a unique edge when it comes to testing big digital reforms. But obviously, this means there’s little room for dissent or public oversight. If the AI makes a bad call or one that benefits the powerful at the expense of ordinary people, there are few mechanisms to challenge or revise it. The result could be a legal system that’s faster, yes — but also more opaque, less accountable, and potentially dangerous in ways that only become clear after the damage is done.
“Law is fundamentally a human endeavor,” said UAE legal expert Ahmad al-Khalil. “Human oversight remains crucial, particularly regarding rights, equity, and interpretation.” He told the told the Khaleej Times, a UAE newspaper, that human oversight is absolutely essential, especially when it comes to something like drafting laws.
For now, the UAE hasn’t revealed which AI tools it’s using or how it plans to make sure they’re safe and accountable. We don’t know exactly how the process will work. But this could end up being a trial that echoes throughout the entire world. The tools already exist. The only question is: how they are used.
