As artificial intelligence continues to reshape the profession, the question of how it could transform government and lawmaking is becoming increasingly urgent. After the UAE recently revealed plans to use AI to assist in legislation, should Australia be following suit?
The United Arab Emirates recently made headlines when Sheikh Mohammed bin Rashid Al Maktoum, Vice President and ruler of Dubai, announced that it will become the first country to use AI not just to assist with legal research or improve efficiency, but to help draft legislation.
This “AI-driven legislative system” will analyse court rulings, executive procedures, and the daily impact of laws – a move described by the UAE government as a “paradigm shift.”
In the UAE, the new Regulatory Intelligence Office will propose law reform and advance legislative processes, as well as tailor “best international policies and legislative practices” to fit the UAE.
“This office will work on creating a comprehensive legislative plan that brings together all federal and local laws in the UAE, connecting them through artificial intelligence with judicial rulings, executive procedures, and public services. The new system will allow us to track the daily impact of laws on our people and economy using large-scale data, and it will regularly suggest updates to our legislation,” the Sheikh said in a post on X.
Today, I chaired the Cabinet meeting at Qasr Al Watan in Abu Dhabi, where we made key decisions to advance legislative processes. We approved the establishment of the Regulatory Intelligence Office within the Cabinet. This office will work on creating a comprehensive legislative… pic.twitter.com/0FbUbxckT8
— HH Sheikh Mohammed (@HHShkMohd) April 14, 2025
Could – and should – this happen in Australia?
Could AI play a meaningful role in shaping the laws of the future in Australia? While adopting AI in a legislative process could play a role in efficiency and accuracy, there are a number of key elements that need to be explored before AI could be used to draft laws.
According to LexisNexis global associations and strategic partnerships director Veronica Rios, AI collaboration needs training data and human supervision at a minimum.
“AI models need a high-quality, accurate, and up-to-date data set for training. Any ambiguous language, bias, or mis-interpretable terms in the training data will affect the model’s outputs, so ensuring the quality of the training set is essential,” she said.
“One effective approach for AI applications involves collaboration between AI and humans. Human legal experts play a crucial role in ensuring the accuracy, clarity, and coherence of legislative outputs. Charles Darwin University has proposed a method called the AI bun, where AI is utilised to complement human legislative input by offering both preliminary and final analysis and review.
“Generative AI, like the tools developed at LexisNexis, uses existing trained content to create new material. However, it is not suited for creating entirely new laws. Instead, these AI tools provide answers, analysis, and summaries based on expert content to help understand and apply current legal positions.”
AI is still yet to be fully regulated – and is still known to hallucinate and, in some cases, be biased.
In 2023, potential bias in AI was flagged in the recruitment and redundancy process, with legal recruiters at the time emphasising that “the human element in the recruitment process will always be crucial”.
Former president of the Law Society of NSW Brett McGrath has also previously advocated for a collaborative approach that leverages existing frameworks and international best practices in regard to AI.
“AI brings with it many challenges and opportunities, and appropriately flexible, scalable and principles-based regulation will help ensure that the community can both reap the benefit of AI and be protected from its potential threat,” he said at the time.
Lawmakers also need to consider the impact of using AI to craft legislation that can change people’s lives, added Rios.
“One of the key pillars of the Rule of Law is transparency of the law. Laws need to be clear and understandable; individuals need to understand the laws they are held to and that they are protected by. Understanding the impact of legislation on people’s lives and rights is key as implications with such high stakes require stricter oversight, supervision and review,” she said.
“AI systems should be transparent. Users must understand how decisions and suggestions are made. The AI systems used for legislative drafting need to provide clear citations supporting drafting decisions.”
In the UAE, the new legislative system will use AI to draft laws in multiple languages, with the aim of making them simple and understandable for all citizens.
UAE solicitor and law drafter Hesham Elrafei also told The Telegraph in the UK that the new system would be “introducing a whole new way” of crafting laws.
“Instead of the traditional parliamentary model – where laws get stuck in endless political debates and take years to pass – this approach is faster, clearer, and based on solving real problems,” he said.
AI tools can also summarise cases where existing laws are discussed or applied, further helping researchers assess the “impact of current legislation and propose future amendments, such as identifying areas of misunderstanding that need clarification”, according to Rios.
While Australia is still using the traditional parliamentary model, Hicksons partner David Fischl said that AI is already assisting in the drafting process – but it may also be slowing down regulatory processes.
“In a democracy such as Australia, our elected representatives openly debate and refine potential laws in Parliament. This rightly takes time to ensure all stakeholder interests are addressed. AI is, no doubt, already assisting drafters and our elected representatives create, review and debate proposed laws – this is a good thing,” he said.
“AI is already assisting stakeholders to influence laws. There’s evidence that AI is also slowing down government regulatory consultation processes and new laws. The US Internal Revenue Service allegedly received 120,000 submissions on a proposed new crypto tax rule. Vested stakeholders made available a website, so that with a few clicks of a button, a detailed AI-generated submission could be generated. If this continues, it’ll be hard and time-consuming for regulators to work through those volumes of submissions without those regulators also using AI.”
While using AI assistance in legislative drafting can lead to a faster legislative process and enable the government to respond to new challenges more efficiently, it’s also important for things like translated content to be reviewed.
“The UAE’s diverse population speaks many languages despite Arabic being the official language. AI tools can aid in translating content, but translations must be reviewed by fluent speakers to ensure accuracy and intent in legislative drafting,” Rios said.
“AI tools [also] have the capability to assist in language, particularly for simplifying legal drafting by identifying sections that require simplification or improvement. Although AI can suggest changes, these modifications must be reviewed by legal experts to ensure that the complexity of certain laws is maintained where necessary and to prevent risks associated with over-simplification.”
Pros and cons of using AI to write laws
The use of AI in lawmaking brings both promising advantages and serious challenges. While AI offers greater speed, precision, data-driven insights, and responsiveness, there remain concerns around transparency, accountability, algorithmic bias, and the potential erosion of human judgment.
“AI models can pull from a broad range of sources to generate legislative analysis on different topics, allowing legislators to research and compare complex legislative topics quickly and more comprehensively. The UAE’s initiative includes looking at global policy research to benchmark UAE legislation against international standards,” Rios said.
“The use of AI in legislative research and drafting will likely accelerate the process, which will make legislative change more responsive and current. Processes like comparing bill drafts against published versions are already possible.”
However, Rios also stressed that there is a risk of over-reliance on AI in legislative drafting due to its “efficiency gains”, as well as potential misinterpretation of legislation if drafted by AI.
“The quality of the AI output is entirely dependent on the quality of the data it is trained and learns from. Any inconsistencies, bias or errors in training data will persist in AI-generated legislation and could make it harder to detect,” she added.
“Fundamental principles regarding legislation in Australia include rigorous debate and policy research. It is important to maintain these principles when using AI for drafting.”
In addition to these concerns, Fischl said that if Australia adopted a similar approach to the UAE, those in Parliament would still be need to address legal loopholes and conduct reviews.
“AI can be used to more rapidly test large numbers of scenarios to find potential loopholes in proposed laws. However, with every change to the drafting of a law to deal with a potential loophole from a scenario, another loophole will emerge. AI will never be able to determine which are the most important scenarios to address in proposed laws,” he said.
“In a democracy such as Australia, our elected representatives must be the ones who make these decisions. AI can and should be used to find trends in data to assist lawmakers. This will improve outcomes for all of us. However, in our wonderful democracy, our elected representatives must still openly debate and make decisions on the final drafting of laws.”
Lauren is a journalist at Lawyers Weekly and graduated with a Bachelor of Journalism from Macleay College. Prior to joining Lawyers Weekly, she worked as a trade journalist for media and travel industry publications and Travel Weekly. Originally born in England, Lauren enjoys trying new bars and restaurants, attending music festivals and travelling. She is also a keen snowboarder and pre-pandemic, spent a season living in a French ski resort.