Abu Dhabi Government and OpenAI have launched Machine-Readable Government (MRG), an initiative described as a new way for authorities to translate a defined set of government laws and policies into machine-readable form to support AI-assisted decisioning.
The programme is led and governed by the Department of Government Enablement – Abu Dhabi (DGE). According to the announcement, MRG takes written policies as inputs, checks a case against those rules, and generates a decision linked to the specific clause used to inform it. The partners say the framework is designed to keep officials involved and to ensure decisions remain traceable to the governing rule.
Officials will retain oversight and accountability, the companies said. The release states that the initiative is not intended to enable fully autonomous decision-making in higher-stakes areas or for determinations that could materially affect individuals’ rights, eligibility, or access to essential services.
Audit trails and human oversight, the partners say
MRG is positioned as an approach that helps standardize routine regulatory compliance assessments, such as those tied to issuing permits and certificates. The announcement claims that, for routine cases, processing can be done autonomously around the clock and return decisions in minutes that otherwise took days of manual review, while maintaining human review for more complex matters.
The release also says that structuring policies into a consistent machine-readable format can help policymakers identify overlaps or potential conflicts across departments. It adds that it may allow teams to test proposed policy changes across related rules before enactment, while keeping existing legislation in place initially.
Pilot in agriculture and food safety
The first pilot use case is described as already in production with the Abu Dhabi Agriculture and Food Safety Authority (ADAFSA), focusing on food imports. The initiative is said to be conducted in collaboration with Inception42, a G42 company.
OpenAI and DGE say the pilot validated a reusable Policy Adoption Engine designed to convert narrative policy into automated compliance checks, with outcomes traceable back to the precise rule that produced them.
Data residency and “fail-closed” security architecture
For AI computation, the partners say processing for MRG is powered by OpenAI’s UAE inference residency. They state that real-time model computation runs on GPUs within UAE data centers under a fail-closed security architecture governed by UAE cybersecurity standards and subject to Abu Dhabi law.
On privacy, the announcement says safeguards are built in from the start and were jointly agreed by DGE and OpenAI. The release also states that the Abu Dhabi government retains full ownership of all data created through the system and that the data is not used to train AI models. It adds that rules govern storage, usage, and deletion of the data.
Expansion planned across government entities
Following the initial rollout, Abu Dhabi says it intends to expand MRG across government entities. Planned applications listed in the release include construction permitting, business and freelance licensing, data access governance, HR compliance, and procurement.
The announcement frames MRG as a pillar of the emirate’s AED 13 billion ($3.5 billion) Abu Dhabi Government Digital Strategy 2025 to 2027, and says it will integrate with TAMM, the emirate’s unified service app that consolidates more than 1,200 public and private services. The partners also claim use cases are active across 40 government entities and that all Abu Dhabi Government civil servants have been trained in AI.
Statements from OpenAI and DGE
OpenAI CEO Sam Altman said the initiative aims to transform public sector delivery and improve responsiveness while “keeping people in control of decision-making.” DGE Chairman Ahmed Tamim Hisham Al Kuttab said the effort focuses on making policy itself “intelligent” so daily processes such as permit approvals and tender decisions can be supported by AI-enabled systems, while maintaining the government’s delivery responsibility.
