The United Arab Emirates (UAE) has made one of the boldest public sector artificial intelligence (AI) commitments globally, announcing a plan to move 50% of government sectors, services and operations to agentic AI within two years. Framed not as incremental digitisation, but as a structural redesign of government itself, the initiative signals a shift from digital government to what could become autonomous government.
“AI is no longer a tool. It analyses, decides, executes and improves in real time. It will become our executive partner to enhance services, accelerate decisions and raise efficiency,” said Sheikh Mohammed bin Rashid Al Maktoum in announcing the initiative.
The announcement is significant not only for its ambition, but also for positioning AI investment as a strategic instrument of national competitiveness. For years, the UAE has invested in the building blocks for this moment, from digital identity infrastructure and smart government services to sovereign cloud capabilities, data strategies and national AI programmes. This latest move appears to take that investment beyond enablement and into operational autonomy.
Agentic AI: The future of UAE’s public sector transformation
At the centre of the strategy is agentic AI, systems capable not only of generating insights but also of taking action autonomously, executing tasks, adapting to changing inputs, and improving performance over time. In a government context, this could span everything from automating case handling and service delivery to supporting policy execution and operational decision-making.
For Manish Ranjan, research director for software and cloud at IDC EMEA, the success of this vision will depend less on raw infrastructure than on whether government institutions can redesign the underlying systems on which AI operates.
“The UAE’s infrastructure capability is very strong, with mature sovereign and public cloud capacity from local hyperscalers and regional providers,” he said. “The compute foundation to support large-scale agentic workloads is largely in place, which few governments globally can claim.”
Ranjan added: “The real determinant of success will be agentic readiness at the data and process layer, not infrastructure. Workflow, policy and process redesign is the hardest part and, in a federal government, a multi-year change management exercise rather than a technology roll-out.”
Agentic AI in action: Redefining government services and policy execution
That distinction is important. Much of the current excitement around agentic AI focuses on the capabilities of the models. Yet for governments, the harder challenge often lies in integrating those systems into fragmented operational environments, aligning them with policy frameworks and ensuring outcomes can be governed at scale.
That concern is echoed by Mohamed Roushdy, CIO at Reem Finance, who described the target as ambitious but credible, particularly given the UAE’s digital maturity.
“The UAE is not starting from scratch,” he said, pointing to mature platforms such as UAE Pass and TAMM, alongside sustained public investment and widespread AI adoption across government entities.
Still, Roushdy explained that major barriers remain. Legacy fragmentation, uneven data readiness, and constraints in sovereign AI compute could slow progress, particularly for sensitive workloads. “Reaching 50% is achievable if defined as AI-assisted or AI-enabled services, particularly for high-volume, low-complexity use cases,” he said. “However, fully autonomous AI decision-making in complex areas remains constrained by trust, governance and accountability challenges.”
Human-in-the-loop frameworks: Balancing automation and accountability
Those questions of trust are likely to define the next phase of public sector AI strategy. As governments move from using AI as a productivity tool to deploying systems that participate in decisions, risk management models will need to evolve.
IDC’s Ranjan argued that public sector leaders should adopt “human-in-the-loop by design” frameworks, establishing which decisions can be fully automated, which require human review and which must remain human-led.
“Bias is a particular concern in multilingual, multicultural populations like the UAE,” he said. “Governments moving to autonomous service delivery must invest in ongoing model auditing, not just pre-deployment testing.”
This reflects a wider shift in how digital trust is being defined. Traditionally, trust in e-government focused on cyber security, privacy and service reliability. In agentic systems, trust increasingly extends to explainability, model oversight and the accountability of machine-led actions.
Setting the benchmark: UAE’s AI leadership in the GCC region
It also raises the larger regional question of whether the UAE is establishing a benchmark that other Gulf Cooperation Council (GCC) members may now feel compelled to match.
According to Ranjan, the answer is likely yes. “For the past decade, the benchmark in GCC government technology has been digital maturity, especially e-service availability and digital identity adoption,” he said. “The UAE is effectively uplifting that benchmark and replacing it with agentic readiness.”
If that holds, the implications go beyond government transformation. It could accelerate investment in sovereign cloud, AI governance platforms, automation software, digital infrastructure and public sector workforce development across the region.
The workforce element is particularly notable in the UAE’s announcement, with every federal employee set to receive AI training. While upskilling is increasingly standard in AI strategies, the scale and mandatory nature of the commitment suggest the government sees talent development as integral to operationalising autonomy.
