UAE publishes design guide for agentic AI as it aims to automate half of government

The United Arab Emirates has released a national design guide for deploying agentic artificial intelligence across public services, part of a stated ambition to automate up to half of government functions. The document sets expectations for building AI systems that can act with autonomy on behalf of public bodies and marks a major step in the country’s modernisation of service delivery.

The publication represents a policy milestone with immediate implications for procurement, regulation and national AI governance. By approving agentic AI for both citizen-facing services and backend processes, the guide is likely to reshape contracting priorities and the capability requirements that public‑sector vendors will need to meet.

Agentic AI refers to systems that can make or carry out decisions and actions without constant human direction. The UAE’s guide focuses attention on how such systems should be designed and governed to support efficiency, consistency and scale in public administration.

Observers say the move will affect a wide range of stakeholders, from ministries and regulators to technology suppliers and citizens. Read coverage of the publication at Global Government Forum:

Key implications and areas for action include:

• Procurement: a shift in contract requirements and supplier evaluation.
• Regulation: new standards and oversight expectations for autonomous systems.
• Public‑sector vendors: updated capabilities and compliance obligations will be needed.
• AI governance: assurance, testing and accountability frameworks will come into focus.
• Citizen services: changes in delivery models, transparency and trust implications.

Practically, public organisations will need to update risk assessments, audit and testing plans, and capacity for oversight. Agencies should consider running controlled pilots before scaling, ensuring human-in-the-loop checkpoints where necessary, and publishing clear explanations of how agentic systems make decisions.

Technology suppliers should prepare technical documentation, compliance materials and interoperability plans so they can bid for new contracts. Regulators will face pressure to define acceptable levels of autonomy, enforcement mechanisms and redress routes for citizens affected by automated decisions.

Call to action: public bodies, vendors and regulators should review the guide, map existing projects against its principles and prepare procurement and governance updates. Early pilots and clear oversight arrangements will be important as the UAE moves to scale agentic AI across government services.

As organisations navigate AI change, Nuvostella helps teams put practical AI systems to work — from generative tools and agents to automation built for real business processes.

Resources & links

• Source report (via Google News)
• Nuvostella Generative Studio