Closing the Transparency Gap: Disability-Inclusive and Ethical AI in Public Services
Posted on November 20, 2026
Director of Research, G3ict
Artificial intelligence is now embedded in the daily operations of public services-from digital identity and telehealth to transportation, benefits administration, and civic participation. For many, these systems promise efficiency and innovation. But for persons with disabilities, they also introduce new risks when accessibility and human rights safeguards are missing. G3ict’s report, Closing the Transparency Gap for Disability-Inclusive and Ethical AI in Public Services, reveals a striking global reality: only 34 of 193 UN Member States publish enough public evidence to show how their public-sector AI systems are governed. As the report notes, “82% of countries... provide no evidence of ethical and accessible AI implementation in public services.” This gap is not theoretical. It affects real people, in real time.
When a telehealth chatbot misinterprets AAC speech, when voice authentication rejects older adults, or when a navigation app sends wheelchair users down inaccessible routes, the harm is immediate. These failures are not simply technical glitches-they are barriers to essential services. As the report’s foreword warns, “An eligibility algorithm or identity check that cannot be examined cannot be challenged.” Without transparency, governments cannot demonstrate that their AI systems uphold disability rights, and persons with disabilities cannot challenge decisions that affect their lives.
Among the 34 countries that do publish evidence, several show strong foundations. Denmark, Estonia, the Netherlands, and Sweden lead the transparency ranking. Yet even these countries reveal gaps-especially in documenting the participation of persons with disabilities in AI governance. Only five countries provide evidence of such participation, despite Article 4.3 of the Convention on the Rights of Persons with Disabilities requiring governments to “closely consult and actively involve” persons with disabilities in decisions that shape their lives.
Countries that publish no evidence are effectively flying blind. AI systems are being deployed faster than governments can monitor them, and without public reporting, accessibility becomes optional rather than required. As the report warns, the pace of deployment “outpaces government officials’ and developers’ capacity to ensure these systems uphold the rights of persons with disabilities.”
Closing this gap begins with three straightforward actions. Governments must publish what they are doing: AI strategies, regulations, accessibility requirements, audits, and outcomes. They must protect the rights of persons with disabilities by enforcing accessibility in procurement, monitoring automated decision-making, and providing clear pathways for redress. And they must participate by involving persons with disabilities in every stage of AI governance, from design to oversight.
Because much of public-sector AI is built by private companies and purchased by governments, procurement is one of the most powerful levers for embedding accessibility and ethical safeguards. Clear procurement requirements can help ensure that accessibility is considered before systems are deployed rather than after barriers emerge. AI will continue to expand across public services. The question is whether that expansion will uphold human rights, protect accessibility, and include the voices of persons with disabilities. Closing the transparency gap is a critical first step toward ensuring that AI serves everyone and that no one is left behind.
For G3ict and its partners, this report is more than a diagnostic. It is a practical tool for advocacy and accountability. It provides governments, civil society organizations, researchers, and industry leaders with evidence they can use to strengthen oversight, improve policy, and advance disability-inclusive AI governance. By identifying both leaders and gaps, the report establishes a baseline against which future progress can be measured.
As governments modernize public services, the choices they make now will determine whether AI becomes a force for inclusion or a new source of exclusion. Transparency is important not simply because it reflects good governance, but because it allows accessibility, ethics, and human rights commitments to be examined, measured, and verified. Public reporting creates the foundation for accountability. Without it, people with disabilities are asked to trust systems they cannot see, understand, or challenge. The report makes clear that this is no longer acceptable in a world where AI increasingly mediates access to essential services.
The path forward requires sustained commitment. Governments must invest in accessible digital public infrastructure, strengthen oversight mechanisms, and ensure that disability-inclusive procurement becomes standard practice rather than an exception. Industry partners must recognize that accessibility is not a feature but a responsibility. The disability community must also be supported not only to participate in AI governance, but to help shape the standards, safeguards, and expectations that define ethical public-sector AI. Ultimately, closing the transparency gap is about more than publishing documents. It is about building public services that people can trust, services that uphold rights, prevent harm, and reflect the lived experiences of the communities they are meant to serve. G3ict will continue working with governments, advocates, and industry to keep accessibility and human rights at the center of AI governance. The opportunity is here. The responsibility is clear. And the time to act is now.

