More than 100 federal officials from across the UAE government attended a workshop this week to begin embedding autonomous AI systems into public sector operations, marking the operational launch of a national program that aims to convert 50% of government services and tasks to agentic AI within two years.
Organized by the National Committee for the Agentic AI Project, the workshop introduced frameworks for classifying government tasks and setting implementation priorities. The effort is being pursued under the UAE Government 4.0 framework, with a focus on using AI to redesign how government work is structured - not simply automate isolated tasks.
The program covers seven tracks: strategy and projects; foresight and strategic intelligence; policies; structures and governance; government performance; global competitiveness; and innovation in government work. Taken together, these pillars suggest the program is positioned as an operating-model change rather than a tech deployment project.
Coordinating across agencies
Federal entities were encouraged to coordinate their AI efforts and reduce duplication. The guiding principle of the workshop was "human leads, AI enables," placing human decision-making at the center while giving AI a larger role in execution and coordination.
"The session marked the operational start of deploying AI models that will bolster future-ready governance," said Huda Al Hashimi, Deputy Minister of Cabinet Affairs for Strategic Affairs.
This session follows the program's first phase in June, when more than 300 officials from 50 federal entities gathered in Dubai. That phase focused on identifying services and operational processes that could be redesigned around agentic AI and launched within 90 days.
The distinction between conventional AI and agentic systems matters. Most government AI deployments to date concentrate on individual use cases such as prediction, analysis, or task automation. Agentic systems are built to operate across multiple steps of a process, coordinating actions and making decisions with limited human intervention.
The real challenge: governance, not technology
For the UAE, the challenge now is less about demonstrating what agents can do than determining where they should be allowed to act, how their performance will be governed, and how responsibilities will be divided between people and machines. The two-year target makes that an organizational design problem as much as a technology one.
For government professionals, this shifts the practical question from "Which tasks can I automate?" to "Which tasks should I automate, and under what rules?" Those questions are not technical issues to be sent to IT departments - they are policy and operational decisions requiring direct input from the people who run services today. Government leaders responsible for AI adoption and regulation may find the AI for government courses and the AI learning path for policy makers directly relevant to this work.
Your membership also unlocks:
