Unlock Editor’s Digest Lock for Free
FT editor Roula Khalaf will select your favorite stories in this weekly newsletter.
The UAE is the most fundamental attempt in the Gulf state to use AI to utilize billions of infused technology, aiming to help write new laws to review and amend existing laws.
The plan for what state media called “AI-led regulations” is further than what can be seen elsewhere, AI researchers said the details are scarce. Other governments are trying to use AI to make them more efficient, from summarizing the bill to improving public service delivery, but they are not actively proposing changes to current laws by calculating government and legal data.
“This new legislative system, powered by artificial intelligence, changes the way laws are written, making the process faster and more accurate,” he said.
Last week, ministers approved the creation of a new cabinet unit, the Regulatory Intelligence Agency, to oversee legislative AI pushes.
Ronnie Medalia, a professor at the Copenhagen Business School, described the plan as “very bold” as the United Arab Emirates appears to have “basically ambitions to turn AI into some kind of co-scholar.”
Abu Dhabi has placed a large bet on AI and last year opened its own dedicated investment tool, MGX. This supported the $30 billion BlackRock AI-Infrastructure Fund, among other investments. MGX has also added AI observers to its own board.
The UAE plans to use AI to track how laws affect the country’s population and economy by creating large databases of federal and local laws and creating public sector data such as court rulings and government services.
AI “will regularly propose updates to our laws,” says Sheikh Mohammad. The government hopes that AI will speed up the law by 70%, according to a read from the Cabinet meeting.
However, researchers noted that it could face many challenges and pitfalls. These range from AI to being inexplicable to users to biases that are triggered by its training data and questions about whether AI interprets the law the same way as humans.
The AI model is impressive, but “they continue to harbor reliability and robustness issues,” warned Vincent Straub, a researcher at Oxford University. “We can’t trust them.”
The UAE’s plan is particularly novel as it involves using AI to predict required legal changes, Straub said. They can potentially save money. The government often pays to law firms to review the law.
“They seem to be taking it a step further. From looking at AI, let’s make it a tool that can help and draft, like an assistant, a tool that can really predict and predict.”
Keegan McBride, a lecturer at the Oxford Internet Institute, said the authoritarian UAE is embracing the digitalization of governments, which employ “easier times” than many democracies. “They can move fast. They can experiment with things.”
There are dozens of small ways governments use AI in law, McBride said, but he hadn’t seen similar plans from other countries. “In terms of ambition,[the United Arab Emirates]is near the summit,” McBride said.
It is unclear which AI systems will use it, and experts said that multiple AIs must be combined.
However, setting guardrails for AI and human supervision is extremely important, the researchers said.
AI can propose something “really, really, really weird” that “makes sense for the machine,” but “it might never make sense to actually implement it in human society,” said Marina de Vos, a computer scientist at the University of Bath.
