Abu Dhabi’s Department of Government Enablement has formally launched Machine-Readable Government, an initiative that translates a defined set of the emirate’s policies and regulations into machine readable form so that artificial intelligence systems can assist experts in navigating complex regulatory decisions. Developed with OpenAI, the project is positioned as a central pillar of the government’s stated ambition to become the world’s first AI-native government by 2027. It also raises a question that will interest oversight-minded observers everywhere: how far can a state push automation of rule application while keeping human office-holders accountable for the outcomes?
The answer the Abu Dhabi Government offers is a framework built around traceability. Under MRG, routine regulatory compliance assessments, such as those tied to issuing permits and certificates, can be carried out consistently, with each decision traceable to the governing rule that produced it. The initiative changes how written rules are applied while leaving the rules themselves unchanged: a case is checked against the encoded policy, and the resulting decision is linked directly to the exact clause that informed it. Routine cases may be processed autonomously, around the clock, returning in minutes decisions that once took days of manual review.
Human oversight, according to the government, remains central throughout. Government teams set the standards, review complex cases, and retain strict responsibility for every final decision. The framework is not intended to make autonomous decisions in high-stakes areas, nor in determinations that materially affect individuals’ rights, eligibility, or access to essential services. That boundary between automated processing and human accountability is the design feature on which the initiative’s credibility with regulators and the public will largely rest.
His Excellency Ahmed Tamim Hisham Al Kuttab, Chairman of the Department of Government Enablement, framed the effort as a restructuring of policy itself rather than a layer of technology on old processes. “When we set out to build the world’s first AI-native government, we knew that applying AI to old processes would only take us so far,” he said. “The harder task was making policy itself intelligent, and that is what a Machine-Readable Government does.” He described the rules that shape daily life, how a permit is approved, how a tender is decided, how a service reaches the public, as now being “as capable as the systems we build to serve our people,” adding that the initiative “puts us on a path no other government has yet taken.”
Sam Altman, Chief Executive Officer of OpenAI, welcomed the collaboration as a model for public sector delivery. “Making AI accessible for everyone requires transforming public sector delivery, helping it become faster, more responsive and more effective,” he said. “By making certain laws machine-readable, while keeping people in control of decision-making, Abu Dhabi has set a new standard for AI-enabled public services.”
The first operational test is already in production. A pilot with the Abu Dhabi Agriculture and Food Safety Authority, conducted in collaboration with Inception42, a G42 company, focuses on food imports and validated a reusable Policy Adoption Engine that converts narrative policy into automated compliance checks, with every outcome traceable back to the precise rule behind it. Expansion is planned across government entities, with upcoming applications in construction permitting, business and freelance licensing, data access governance, HR compliance, and procurement.
Meanwhile, beyond processing speed, the structuring of policies into a single machine-readable format gives policymakers an oversight tool of their own: a clear view of where regulations overlap or conflict across departments. Officials can test proposed policy changes against related rules before enactment, helping simplify and strengthen regulation while preserving established legislation.
Data governance arrangements are central to the initiative’s accountability framework. AI processing for MRG runs on OpenAI’s UAE inference residency, meaning real-time model computation occurs strictly on GPUs within UAE data centers under a “fail-closed” security architecture governed by UAE cybersecurity standards and subject to Abu Dhabi law. This makes the UAE, according to the announcement, the first country in the Middle East to deploy OpenAI’s GPT models using in-country inference infrastructure, spanning everything from internal government research to automated compliance checks on commercial applications.
Privacy protections were built in from the start, with safeguards jointly agreed by DGE and OpenAI. The Abu Dhabi government retains full ownership of all data created through the system; that data is not used to train AI models, and clear rules govern its storage, use, and deletion.
MRG sits within the AED 13 billion (approximately $3.5 billion) Abu Dhabi Government Digital Strategy 2025 to 2027 and will integrate with TAMM, the emirate’s unified service app consolidating more than 1,200 public and private services. With use cases active across 40 government entities and all Abu Dhabi Government civil servants trained in AI, the initiative is presented as the foundation that makes government policy itself ready for artificial intelligence, and, in doing so, a live test of how a government embeds accountability into automated administration. Whether that test succeeds will depend on how the line between autonomous processing and human responsibility holds as the system scales.