Half of UAE federal government operations are set to shift to agentic AI models within two years. Officials are now working to define which tasks AI can handle autonomously, raising new questions about accountability and oversight.
The United Arab Emirates is accelerating its push into artificial intelligence, launching a national initiative to convert 50% of federal government operations, services, and tasks to agentic AI models within two years. This month, over 100 federal officials gathered for a workshop led by the National Committee for the Agentic AI Project, marking the operational start of the program. The session focused on setting priorities, timelines, and-most critically-developing frameworks to determine which government tasks can be fully delegated to AI agents.
Since publishing its national AI strategy in 2017 and appointing Omar Sultan Al Olama as the world’s first minister of state for artificial intelligence, the UAE has built digital identity, sovereign cloud, and data-sharing infrastructure that many governments are still working to implement. Now, the country faces a new challenge: drawing a clear line between tasks that AI can execute autonomously and those that require human oversight. Huda Al Hashimi, Deputy Minister of Cabinet Affairs for Strategic Affairs, described the workshop as the practical beginning of deploying AI models across government functions. The program is structured around seven pillars, including strategy, foresight, governance, performance, competitiveness, and innovation, with officials also discussing how to avoid redundant deployments across agencies.
The project is guided by the principle “human leads, AI enables.” Yet, there is a tension between the founding vision-autonomous execution of government operations-and the implementation language, which emphasizes human primacy. This gap is precisely what the new classification frameworks aim to address. Until these frameworks are published, questions remain about accountability: what recourse do citizens have if an AI system mishandles their case, and where does liability fall? These are issues no government has yet resolved at scale, but the UAE is now on a two-year clock to provide answers.
Oversight for the initiative rests with Sheikh Mansour bin Zayed Al Nahyan, with a taskforce chaired by Mohammad Al Gergawi, Minister of Cabinet Affairs. In May, the Cabinet approved the first set of government services to use agentic AI tools, alongside the largest training program in UAE government history-80,000 federal employees, from ministers to new hires, will participate. The same session saw the passage of a national AI healthcare policy and new frameworks mandating digital records as the official source of government data. Training events in June drew hundreds of participants from dozens of federal institutions, all working to identify which services are suitable for agentic AI within 90 days.
Governments in Asia and Europe are watching closely, most still lacking concrete targets or timelines for AI adoption. The UAE’s commitment to a specific number and deadline means it will soon provide the first large-scale evidence of how agentic AI impacts public service delivery, staffing, and error rates. As seen in other sectors-such as the challenges AI agents face with B2B website pricing, explored in this related analysis-the real test lies in defining which decisions can be safely automated. The forthcoming task classification frameworks will be the first attempt by any government to formally document which decisions it is willing to entrust to machines.
According to official UAE government data, the country’s digital transformation efforts have already resulted in more than 150 million data points being tracked monthly by AI-powered performance systems. The UAE federal government employs over 100,000 people, and the new agentic AI initiative is expected to impact a significant portion of these roles as automation expands. The government’s AI strategy, launched in 2017, positioned the UAE as a regional leader in digital governance, with ongoing investments in cloud infrastructure and data security to support its ambitious targets.
