Abu Dhabi’s Government Digital Strategy 2025–2027 allocates AED 13 billion to a programme targeting more than 200 AI solutions and full sovereign cloud adoption for government operations. These targets put AI at the centre of how the emirate plans to run public services. [1]
Across the Gulf Cooperation Council (GCC), governments are approaching the same challenge from different starting points. Some are deploying AI across everyday work. Others are building the policies, data standards and skills needed to use it responsibly. Together, these developments show how closely AI adoption is becoming tied to the organisation of government.
What AI native government means
AI-native government describes an operating model in which AI is built into how public services work. A system might help process documents, anticipate a service need or guide a civil servant through a complex request. Reliable data, clear responsibilities and regular checks on performance are part of the design.
AI governance provides the rules around that model: what a system may do, who checks its output and who remains responsible. Answering a question about a licence is different from deciding whether to issue one. The degree of human oversight should reflect the consequences of the task.
The OECD’s 2025 research offers useful context. Of 200 government AI use cases it examined, 57% supported the automation, streamlining or tailoring of processes and services. The report also identifies barriers in skills, data access and impact measurement. Those themes recur in the GCC initiatives behind the following five trends. [2]
1 Better data for better services
A public service can only give a reliable answer if it can find the right information. An outdated record or conflicting guidance can undermine even a capable AI system. This makes data quality a service issue as well as a technical one.
Abu Dhabi’s strategy links AI adoption to sovereign cloud infrastructure and a shared enterprise platform. Dubai’s updated Data Manual, launched in July 2026, addresses the information that these systems rely on. It sets out standards for data quality, ownership, sharing and compliance, with a stronger focus on preparing data for AI. [1] [3]
For agencies, the practical work starts with knowing which records are authoritative, who maintains them and how changes reach connected services. Cloud location is another consideration, but keeping data within a jurisdiction does not resolve inconsistent records or excessive access permissions. AI needs both dependable information and clear rules for using it.
2 Clearer rules for everyday AI use
AI systems can change as their data and models change. Governance therefore needs to continue after launch. In July 2026, Saudi Arabia’s national portal reported the launch of the Saudi Data and Artificial Intelligence Authority’s National AI Risk Management Framework. It covers risk assessment and treatment, followed by continuous monitoring and review. [4]
Oman’s safe and ethical AI policy takes a similar lifecycle view. Its May 2025 announcement describes performance assessments, documented decisions and human oversight in sensitive situations, with requirements for entities developing or using AI in both the public and private sectors. [5]
The status of these documents varies across the region. Bahrain’s General AI Policy requires compliance by government entities and preserves human control over important decisions. Qatar’s ethical-use guidelines explicitly describe themselves as non-binding guidance. Each deployment needs to reflect the rules that apply to its country and service. [6] [7]
A useful starting point is to give each service an accountable owner, define what the AI is allowed to do and establish when staff must intervene. Those arrangements should be reviewed whenever the system’s role changes.
3 More proactive public services
Connected services create opportunities to anticipate what a person needs next. In October 2025, Abu Dhabi’s Department of Government Enablement (DGE) reported that TAMM brought together more than 1,100 services across 90 public and private entities. It described proactive support for tasks such as identity renewals and health appointments. [8]
AI agents can also carry out sequences of tasks within a defined workflow. Dubai Electricity and Water Authority (DEWA) reported in July 2026 that it had embedded agents in digital design and testing, including generating test plans and executing tests. These examples show how AI can support the work behind public-facing services. [9]
As systems take on more actions, their limits need to be explicit. An assistant could help someone prepare a licence application, check for missing documents and route it for review. Permission to approve the application would be a separate decision. Keeping a record of the system’s actions and providing access to human review become essential as its role expands.
4 Public teams learning to work with AI
Large deployments are bringing AI into the daily work of civil servants. In July 2026, DGE reported a Microsoft 365 Copilot rollout to 26,000 employees across 27 entities, alongside 9,000 existing licences. The programme includes training, certification and readiness assessments, recognising that access to a tool is only one part of adoption. [10]
Qatar’s GovAI programme focuses on how agencies move from an idea to implementation. In December 2025, the Ministry of Communications and Information Technology described a common process for selecting use cases, assessing feasibility and developing solutions, with governance built in. [11]
Kuwait is also examining the capabilities needed for responsible adoption. Its September 2026 work with the United Nations Development Programme and the Central Agency for Information Technology assesses the AI ecosystem, government use, and regulation and ethics to inform future priorities. [12]
For public teams, this means learning how to evaluate outputs, handle exceptions and assess suppliers. Managers need to identify where AI helps and where staff judgement remains essential. These capabilities allow an agency to keep improving a service after the initial rollout.
5 Measuring results in public services
The UAE Ministry of Finance offers an example of measuring AI-enabled services. For the first half of 2026, it reported 97.11% first-contact resolution, an eight-second average speed of answer and 95.43% customer happiness at its call centre. The service combines generative AI with speech-to-text, sentiment analysis and other platforms. [13]
These figures describe the performance of the whole service. They do not isolate how much of the result came from AI, staffing or process changes. That distinction matters when deciding whether to expand an initiative.
A stronger evaluation would compare performance before and after deployment, tracking the time and cost of resolving a complete request alongside accuracy and repeat contacts. It should also examine when staff need to correct an answer and whether users can challenge a decision. Testing relevant languages and accessibility needs helps reveal who benefits and where the service still falls short.
The next step for GCC governments
The GCC shares some reference points, including the regional AI ethics manual that Bahrain adopted alongside its national policy in July 2025. Delivery still takes place through distinct national and local institutions, with different rules and levels of readiness. [14]
The five trends point to a common priority: making AI a dependable part of public service. Progress becomes clearer when an agency can show how a service has improved, explain the role AI played and identify who is responsible when something goes wrong. Reliable data and capable teams make that possible; regular evaluation shows where to go next.
Sources
[1] Abu Dhabi Department of Government Enablement. Abu Dhabi Government Digital Strategy 2025–2027. 2025.
[2] OECD. Governing with Artificial Intelligence. 18 September 2025.
[3] Digital Dubai. Launch of the updated Dubai Data Manual. 1 July 2026.
[4] Saudi national portal and Saudi Press Agency. SDAIA introduces national framework for managing AI risks. 14 July 2026.
[5] Oman Ministry of Transport Communications and Information Technology. National Policy for the Safe and Ethical Use of Artificial Intelligence Systems. Announcement 18 May 2025.
[6] Bahrain Information and eGovernment Authority. General Policy for the Use of Artificial Intelligence. Approved 20 May 2025; sections 3, 4 and 8.
[7] Qatar Ministry of Communications and Information Technology. Principles and Guidelines for Ethical Use of Artificial Intelligence. November 2024; p. 2.
[8] Abu Dhabi Department of Government Enablement. TAMM services and integration across entities. 10 October 2025.
[9] Dubai Electricity and Water Authority. Agentic AI deployment across digital platforms. 6 July 2026.
[10] Abu Dhabi Department of Government Enablement. Microsoft 365 Copilot rollout and Frontier Employee Programme. 6 July 2026.
[11] Qatar News Agency reporting MCIT. GovAI programme objectives and future ambitions. 9 December 2025.
[12] United Nations Development Programme. Kuwait Assesses AI Opportunities and Priorities. 9 September 2026; activity dated 8 September.
[13] UAE Ministry of Finance. Generative AI call centre and H1 2026 operational performance. 23 July 2026.
[14] Bahrain Information and eGovernment Authority. National AI policy and adoption of GCC ethics manual. 27 July 2025.