The United Arab Emirates plans to use artificial intelligence agents across half of its government operations within two years. The move aims to make public services faster and more efficient. However, it also raises questions about how the country will monitor these systems and keep them safe.
Omar Sultan Al Olama, the UAE’s minister of state for artificial intelligence, outlined the country’s plans on October 14. He spoke at the World Economic Forum’s Annual Meeting of Global Future Leaders in Dubai.
Al Olama said the government uses a special oversight system known as a control tower agent. The country also checks AI activity through both automated audits and human reviews. These steps aim to make sure that AI agents act within set limits.
He said safety and trust are vital to the plan. The government wants to use AI to improve daily life without putting citizens at risk.
“There is an oversight that ensures that we understand what agents are doing,” Al Olama told the audience. He added that the checks are ongoing and involve both AI systems and people.
UAE Sets Two-Year AI Target
In April, the UAE Cabinet announced a plan to move 50% of federal government sectors, services, and operations to agentic AI within two years.
Agentic AI refers to systems that can carry out tasks and take action with limited human input. Unlike tools that only answer questions, these systems can follow steps, make choices within set rules, and complete work.
The UAE wants to redesign public policies and work processes around these systems. The goal is to cut costs, speed up services, and help government offices act before problems grow.
Under the plan, AI systems could take on more routine work. This may help public bodies use staff time and other resources more effectively. Yet, the scale of the project also makes oversight a major concern.
Al Olama’s comments show that the UAE is not treating AI adoption as a simple technology upgrade. The government also wants systems that can be checked and held to clear rules.
Human Checks Remain Important
The UAE’s control tower agent is meant to help officials track what other AI agents do. The system supports ongoing checks on their actions and decisions.
The government also uses AI agents to audit other agents. Human reviewers provide another layer of oversight. This approach aims to catch mistakes and reduce the risk of harmful actions.
Still, monitoring large numbers of AI agents may be difficult. Systems can act quickly and may carry out several tasks at once. Clear rules, reliable records, and human review can help officials spot errors and respond to problems.
The challenge may grow as AI agents gain access to more government systems and services. A mistake in a low-risk task may be easy to fix. An error linked to public records or key services could have wider effects.
The UAE has not shared every detail of how its control tower works. Further information about its rules, review process, and response to failures will help show how the plan will operate in practice.
AI Use Grows Across the UAE
The UAE’s push into agentic AI follows a sharp rise in public use of AI tools. Microsoft reported that more than 73% of the UAE population used AI during the second quarter of 2026.
The figure measures the share of adults who used a generative AI product during a set period. By comparison, global AI use reached 19% in the same quarter, according to Microsoft’s report.
The figures show that AI tools have become common in the UAE. The government’s plan seeks to take this use further by placing AI agents inside public services and operations.
However, wider AI use also brings new security risks. The UAE and other countries in the region have faced rising cyberattacks in recent years. Some attacks use AI to automate tasks or improve the speed and scale of operations.
Political tensions and armed conflicts in the Middle East add to these concerns. Government systems may hold sensitive data and support services that people rely on every day.
Security Risks Could Shape the Plan
As the UAE expands its use of AI agents, officials will need to balance speed with safety. Strong access controls, regular testing, and clear limits on what agents can do may help reduce risks.
Human oversight will also matter when an agent makes an unusual choice or encounters a task outside its role. Officials need ways to stop systems, review actions, and correct errors when needed.
The UAE’s two-year target is ambitious. Its success will depend not only on how many government services use AI, but also on how well those systems are managed.
Al Olama’s remarks underline the government’s view that trust must grow alongside automation. As more public work moves to AI agents, ongoing audits and human checks are likely to remain central to the UAE’s approach.
