UAE targets 50% federal government automation via agentic AI within two years

The United Arab Emirates has launched a strategic national programme to transition 50% of its federal government operations, services, and tasks to agentic AI models within two years. This initiative marks the first large-scale attempt by a sovereign state to formally define the operational boundaries of machine autonomy in public administration.

To achieve this target, the UAE has entered an operational phase centered on workshops for federal officials. These sessions establish classification frameworks that distinguish between tasks capable of being handled autonomously by AI and those that strictly require human decision-making. The project is underpinned by the guiding principle that "human leads, AI enables."

A systemic overhaul

The scale of the transition is supported by a massive human capital investment, including a training programme for 80,000 federal employees, ranging from cabinet ministers to new hires. The programme is organized around seven strategic pillars: strategy and projects, foresight and strategic intelligence, policies, structures and governance, government performance, global competitiveness, and innovation in government work.

This push is not starting from zero. The UAE already operates a proactive AI-powered performance system that tracks more than 150 million data points every month, providing a data-rich foundation for the deployment of agentic models—systems capable of taking independent actions rather than merely offering recommendations.

The shift to machine agency

While many nations have published broad AI strategies, the UAE is the first to attach a specific quantitative target and a strict deadline to the deployment of autonomous AI. The transition from generative AI, which assists humans, to agentic AI, which executes tasks, represents a fundamental shift in how a state interacts with its citizens and manages its internal bureaucracy.

By creating formal classification frameworks, the UAE is effectively building a blueprint for the legal and ethical boundaries of machine agency. This approach addresses critical questions regarding liability and citizen recourse when an autonomous system makes an error, providing a real-world case study for other governments considering similar paths.

The road to autonomy

This initiative is the latest step in a trajectory that began in 2017, when the UAE appointed the world's first minister of state for AI, Omar Sultan Al Olama. Having already established sovereign cloud and digital identity layers, the government is now moving toward a fully digitized, autonomous administrative state.

Observers will now be watching how the UAE implements these classification frameworks in practice. The primary challenge remains the tension between the efficiency of autonomous execution and the necessity of human oversight in sensitive governance decisions. Whether the 50% target can be met without compromising accountability will likely determine the viability of agentic AI for other sovereign entities.