AI to reboot UAE’s skyscraping tech ambitions, away from fossil fuels

This novel “AI-driven regulation” system, to be overseen by the newly-created Regulatory Intelligence Office, a unit tasked with managing the AI’s integration into its lawmaking process that will change how laws are created, making the process faster and more precise, according to Sheikh Mohammad bin Rashid Al Maktoum, the UAE’s Vice President, Prime Minister, and the ruler of Dubai.

The AI system, announced just weeks ago, is expected to comb through a comprehensive database of federal laws, court rulings, and government services, and come up proactively with legal updates.

AI’s role, said UAE officials attending the third edition of the Machines Can See Summit in Dubai over the weekend — touted as the largest computer vision event and a leading AI gathering in the Gulf region — goes beyond just drafting; the system will track the real-world impact of new laws on the Emirates’ population and economy, allowing the administration to adapt legislation in real time. The plan, though, has raised concerns among researchers and technologists, who have flagged the propensity of AI models to produce unreliable outputs, hallucinate and its potential shortcomings with legal interpretations.

These risks notwithstanding, the UAE has made AI a national priority, pumping in billions into the sector through its new sovereign wealth vehicle, MGX. While its fascination with tech is not new, it is the scale of the Emirates’ plan, alongside that of neighbouring Saudi Arabia, that is key. The UAE’s centralised government structure is a facilitator of this kind of rapid technological experimentation, something that could be difficult to achieve in other countries.

“Humans usually like to speak, to discuss, to get options and also to negotiate. And this is what the machine is (now) doing. We are going to a more human-centric approach. In Dubai, we are designing a multi AI agent, where a single AI can talk to one or multi AI agent in order either to perform tasks or execute commands or get a notification in order to finish a service. So what we’ll see in the future, and I think this is one of the most complicated things that we’ll see, we’ll see more AI working with the digital tool. So each one of us will have a digital tool,” Hamad Obaid Al Mansoori, Director General of Digital Dubai, said at the first day of the Machines Can See summit here on April 23.

Mansoori said the provincial administration is gearing for a future where AI agents or virtual ‘Metaverse-type’ avatars of citizens could be potentially designated to carry out tasks on behalf of citizens, such as applying for a licence or a business permit. That is the level of crystal ball gazing that the UAE administration is willing to do in its pursuit of AI leadership.

Within the Emirates, Dubai has taken a lead in AI, and policymakers hope to keep it that way, Omar Sultan Al Olama, head of the UAE Ministry of State for Artificial Intelligence, Digital Economy and Remote Work Applications, said at the event that the administration is following a whole-of-government approach to keep that lead. “There is a saying… Every morning in Africa, a gazelle wakes up. It knows it must run faster than the fastest lion or it will be killed. Every morning a lion wakes up. It knows it must outrun the slowest gazelle or it will starve to death. It doesn’t matter who you are, when the sun comes up in Dubai, you’d better be running,” Olama said in his opening keynote.

The approach is backed by action. In March 2024, the UAE floated MGX, a tech-investment company by two state-owned entities – global investment fund Mubadala and an artificial intelligence firm G42 — with a target size of $100 billion that will invest in AI infrastructure, including data centres and chip-making facilities. It has also set up a $10 billion AI venture-capital fund.

Earlier in 2023, the Technology Innovation Institute, a state-owned research body, unveiled Falcon, a large language model or LLM, and is currently working on new, smaller models. As part of these efforts, G42 had unveiled Nanda, a Hindi LLM while Mubadala had earlier invested in Anthropic, a top global AI startup.

Countries such as India have been slower to get started. On Saturday, the Indian government said it had selected a Bengaluru-based start-up Sarvam to build the country’s first indigenous AI large language model. The start-up, chosen from among 67 applicants, will receive support from the Union government in terms of compute resources to build the model from scratch and is the first to get approved for sops under the ambitious Rs 10,370 crore IndiaAI Mission to build its own LLM model.

In the Emirates, non-AI companies are simultaneously being roped in to prep up the region’s infrastructure to match its AI projections. E&, an Emirati telecom company, has been roped in as part of a project to build a 45,000-km long undersea cable that makes its way around south Asia, Africa to the UK. Khazna, an arm of G42, is also building peripheral infrastructure, all part of a concerted plan.

Last August Mubadala invested in Yondr, an American data-centre developer. The UAE is also reported to be in talks with TSMC, the world’s largest chipmaker, to build a semiconductor foundry in the Emirates.

A big concern, though, is the worsening relation between America and China, analysts attending the conference told The Indian Express. Companies from the UAE are reliant on American companies for technological and capital support, even as most of them have heavy exposure to multiple Chinese players for hardware and software requirements. Having to pick sides would be an emergent risk for the tech-focused administrators sitting in Abu Dhabi, as will be the case for those in other capitals around the world.

(The writer was in Dubai at the invitation of the Polynome Group, which organised the MCS 2025 in partnership with the UAE government)