Building Accountable Artificial Intelligence in Government: A Practical Reform Agenda

Artificial Intelligence (AI) is moving fast. Investment is surging, governments are rolling out systems across public services, and new AI models are being released with increasing power and sophistication. The risk is not only that rules and regulations arrive late. But that AI becomes embedded in the state in an opaque and unaccountable way that is almost impossible to unwind.

Ahead of the India AI Impact Summit this week, the Open Government PartnershipThe Open Government Partnership (OGP) is a multi-stakeholder initiative focused on improving government transparency, ensuring opportunities for citizen participation in public matters, and strengthen... More (OGP) is making a clear case: AI governance is not just a set of principles, it is a necessary set of practical reforms that countries can co-create and implement now.

The gap: global principles, uneven implementation

Over the last few years, there has been significant progress on international frameworks for responsible AI, thanks to recommendations from OECD and UNESCO. But many countries still require turning high-level commitments into working mechanisms. This looks like clear disclosure rules, enforceable safeguards in procurement, independent oversight, and meaningful public participationGiving citizens opportunities to provide input into government decision-making leads to more effective governance, improved public service delivery, and more equitable outcomes. Technical specificatio....

The consequences of getting this wrong are real. High-profile failures, from flawed fraud detection systems to bias in facial recognition, show what happens when automated systems are deployed without robust safeguards and accountability.

OGP’s proposition: move from norms to concrete reforms

OGP is a partnership designed to help countries translate open government values into domestic action. We do not create standards. But through OGP commitments and the Open Gov Challenge, national and local governments work with civil society to co-create reforms that are time-bound, measurable, and independently assessed.

OGP’s model maps directly onto what AI governance needs most right now:

• TransparencyAccording to OGP’s Articles of Governance, transparency occurs when “government-held information (including on activities and decisions) is open, comprehensive, timely, freely available to the pub... More: the public knows where AI is used, for what purpose, with what data, and the responsible stakeholders.
• Participation and inclusionOGP participating governments are working to create governments that truly serve all people. Commitments in this area may address persons with disabilities, women and girls, lesbian, gay, bisexual, tr... More: affected communities and civil society shape the rules and choices, not only technical experts or vendors.
• Accountability and oversight: checks on deployment, monitoring of impact, and routes for remedy.
• Human rightsAn essential part of open government includes protecting the sacred freedoms and rights of all citizens, including the most vulnerable groups, and holding those who violate human rights accountable. T..., diversity, and equity: AI does not amplify discrimination, undermine privacy, or erode trust in public institutions.

Proof it can be done: reforms already underway in OGP countries

Across the world, reformers in government and civil society are already using OGP to co-create and implement practical measures to ensure AI is safe, ethical and public purpose driven.

Examples include public sector AI registers, published impact assessments for automated decision-makingAs governments are increasingly turning to algorithms to support decision-making for public services, transparency and accountability of algorithm use is necessary to minimize their harm and increase ... More, multi-stakeholder governance mechanisms, and dedicated oversight bodies. 

Reforms span a wide range of contexts, including:

• Canada: agencies using automated decision-making must conduct and publish impact assessments prior to deployment.
• Colombia: the development of a multi-stakeholder dialogue mechanism to design AI governance.
• Kenya: the establishment of a framework for inclusive, safe, and responsible governance of emerging technology, focusing on AI.
• Nigeria: produced an Open Gov Challenge Award-winning commitmentOGP commitments are promises for reform co-created by governments and civil society and submitted as part of an action plan. Commitments typically include a description of the problem, concrete action... to strengthen the legal framework for digital governanceAs evolving technologies present new opportunities for governments and citizens to advance openness and accountability, OGP participating governments are working to create policies that deal with the ... More through establishing a National Multi-Stakeholder Coordination Mechanism for digital transformationGovernments are working to increase access to and quality of government services, improve transparency, and create opportunities for participation by using information and communications technologies.... More and developing a National Policy on Ethical, Safe and Responsible Use of Digital Technologies.
• Uruguay: established a public oversight body for government AI use.

This represents a shift from closed technical adoption of AI systems in the public sector to publicly accountable and participatory AI governance.

A practical global offer: peer learning that supports implementers

This week’s India AI Impact Summit will include many conversations about safety, innovation, and competitiveness. Those questions matter. But public sector AI governance also needs a different kind of infrastructure: global networks where implementers can learn from each other, share templates, and solve real delivery problems.

That is why OGP is investing in the Open Algorithms Network. It convenes government practitioners, civil society, and experts working on algorithmic transparency and accountability, with a focus on practical implementation support. The network includes early “open algorithms” reformers, as well as newer actors shaping AI governance through OGP commitments.

A message of urgency 

The next two years are a critical window of opportunity, but one that is closing fast. Once AI systems are procured, vendors are locked in, and institutions settle into routines. Democratic governance becomes harder to retrofit. Countries that build openness and accountability from the start will avoid costly course corrections, build public trust and harvest the rewards of well-deployed AI systems.

OGP’s message going into the India AI Impact Summit is straightforward: AI governance should be built with the public. By default, this will then produce the infrastructure for the public.  Countries do not need to wait for perfect global alignment to begin. They can start by adopting concrete reforms that make AI use in government open and accountable, building stronger norms over time.

OGP stands ready to work with reformers in government and civil society to turn global principles into country-owned action plans, backed by independent reporting and grounded in human rights.