As artificial intelligence (AI) continues to evolve, it is increasingly becoming a valuable tool for electoral authorities worldwide. While many applications remain low-risk, a growing number of electoral management bodies (EMBs) are beginning to explore advanced AI technologies. These integrations aim to enhance election administration throughout the electoral cycle and adapt operations for the AI era. Such innovations are not simply added to existing processes; they fundamentally reshape core election frameworks and communication methods, promising improved voter services and solutions to complex administrative challenges. However, the incorporation of AI also introduces new risks, particularly in essential processes like voter roll management, vote counting, and voter outreach. Concerns range from potential biases and inaccuracies in AI algorithms to issues of uneven institutional adaptation due to varying levels of AI literacy.
As EMBs deepen their AI integration, these risks transition from theoretical to real-world challenges. In response, many EMBs are reassessing their strategies and updating policies to ensure that AI improves electoral administration without undermining democratic values. However, because the use of AI in elections is still largely uncharted territory, there are few precedents or best practices. As a result, each EMB’s approach to adopting and regulating AI technologies is unique. To capture and analyze these varied institutional shifts, International IDEA has conducted a global survey of EMBs to document their diverse experiences with AI as an electoral technology.
In partnership with Microsoft, Arizona State University (ASU), and The Elections Group (TEG), International IDEA has launched the AI + Elections Clinic Skills Hub as part of the Mechanics of Democracy Lab hosted by ASU. This hub is envisioned as a central repository for best practices and use cases, empowering electoral actors to responsibly navigate newly emerging AI-driven election technologies.
The Skills Hub includes case studies created by International IDEA, informed by interviews with electoral authorities. These case studies cover the unique AI applications and governance frameworks in countries such as Estonia, Mexico, the Philippines, Pakistan, South Africa, Albania, Nigeria, the UK, and Australia. Together with other resources available through the Clinic, these studies serve as valuable capacity-building tools for EMBs aiming to harness AI innovations. The goal is to reduce administrative burdens, enhance electoral processes, and redefine the voter experience, all while upholding democratic principles. These case studies offer critical insights into how EMBs can prepare for AI integration while avoiding common pitfalls, guiding the development of internal procedures, staff training, organizational structures, and AI policies.
The Skills Hub builds on International IDEA’s extensive work regarding AI in electoral governance, including the Artificial Intelligence for Electoral Management report, which delves into the opportunities and challenges of AI in electoral management. This initiative also encompasses the AI for Electoral Actors Workshop series conducted across five countries, aimed at enhancing institutional capacity for thoughtful AI use in elections. This website represents a pivotal step in understanding the practical experiences of EMBs with AI integration.
While EMBs may share similarities in their AI usage and governance models, each case presents distinct insights into the tangible realities of AI implementation. The experiences of Albania and Australia exemplify how robust governance frameworks and real-world testing are essential for the responsible deployment of AI technology.
Albania
The Central Electoral Commission (CEC) of Albania adopts an empirical approach when integrating new technologies into its operations. Following this philosophy, the CEC implements a rigorous testing protocol for all new AI technologies, requiring meticulous prototyping and evaluation with strict human oversight before any widespread deployment. During the 2025 partial mayoral elections, the CEC trialed a new AI-driven image analysis tool for vote tabulation, marking its first practical use in an election context. The CEC’s experience serves as a valuable case study in the significance of real-world pilot testing.
The image analysis tool scanned ballots processed by polling station staff to expedite vote counting and results transmission. Since this was the tool’s initial legally mandated pilot test, it was only implemented in 3% of polling locations, with manual counting continuing in parallel. Initially, the system demonstrated strong performance and consistent accuracy; however, oversight staff encountered two unexpected issues due to the physical conditions in the polling booths. Firstly, the system occasionally counted votes twice if ballots were held up for too brief a moment. This problem became evident as the night progressed, with fatigued poll workers handling ballots too quickly for the system to function properly. Secondly, the AI struggled to recognize ballots with minor inconsistencies, leading to valid votes frequently not being counted. This challenge became particularly pronounced due to the wear and tear from regular handling, causing ballot markings to fade and diminishing the system’s reliability over time.
This example underscores an essential lesson: many potential risks associated with new technologies are only evident once they are practically implemented. Although preliminary evaluations and risk assessments play a crucial role in developing AI tools, unanticipated conditions can arise in the field that substantially affect the validity and reliability of the systems. For the CEC, this experience emphasized the necessity of thorough prototyping and real-world testing to identify and address risks before they jeopardize election integrity.
Australia
The Australian Electoral Commission (AEC) takes a cautiously conservative approach to AI integration, ensuring that core electoral administration and voting processes remain strictly analog. While the AEC maintains rigorous standards to protect the integrity of its core functions, it is exploring AI tools to boost administrative efficiency, internal productivity, communication, and voter services. Central to these initiatives are not only the technological aspects of application development but also a careful examination of how institutional design and regulation must evolve in tandem to establish adequate support and safety measures for new AI tools.
To facilitate this careful approach, the AEC is formulating internal AI guidelines and strategies in alignment with the Australian government’s efforts to establish a comprehensive governance framework for AI in the public sector. This multilayered system, which the AEC must navigate while considering new AI applications, encompasses governance frameworks, assurance mechanisms, ethical principles, and technical standards. These governance instruments serve to outline the parameters for AI usage while providing resources that enable public agencies to build their capacity for AI integration, including training programs and a government-sanctioned sandbox environment for testing AI tools without reliance on third-party services. This addresses the reality that many public agencies lack the necessary frameworks, standards, resources, and expertise required for effective AI project execution.
Alongside the government-wide AI policy, the AEC has established internal support structures to navigate this complex governance landscape and promote responsible AI tool development. The Commission has appointed two government-mandated supervisory roles: the Chief AI Officer, who spearheads AI development, and the Chief AI Accountable Officer, who oversees the responsible use of AI across the AEC. Additionally, the AEC has created an AI working group—a dedicated forum open to all staff that focuses on discussing, developing, and testing ideas for new AI applications. This platform fosters internal transparency, guiding staff on AI procurement and enhancing awareness and literacy around AI, ensuring a participatory adoption process.
Transparency serves as the core accountability measure in this system. Commonwealth agencies are required to publish regularly updated transparency statements that explicitly outline which AI tools they use, their intended purposes, and the conditions surrounding their deployment. By choosing to be open to public scrutiny, the AEC mitigates the risk of misaligned governance and provides an additional layer of safeguards against problematic AI adoption. This approach illustrates how a layered policy infrastructure—rather than a singular regulation—can strengthen AI accountability and implementation by ensuring that all stakeholders, including the public, are considered in the development of new applications.
Why This Work Continues to be Important
The experiences of the nine EMBs showcased in the AI + Elections Clinic Skills Hub represent only the beginning of a much larger trend. In just a few years, the AI use cases highlighted in these case studies have transitioned from theoretical concepts to practical applications. As AI technologies become increasingly sophisticated, the uptake of AI in electoral management will likely follow suit. However, EMBs are acutely aware of the risks that accompany AI use in elections, ranging from systemic errors to discrimination and disenfranchisement. This understanding is reflected in the growing trend toward developing comprehensive AI strategies that establish clear norms around the usage of AI in electoral management while exploring innovative ways to enhance various aspects of EMB operations.
If risks are not effectively managed through institutional adaptation, the integration of advanced AI functionalities into core electoral activities may not only cede control over critical processes but also obscure the decision-making mechanisms behind AI-generated outcomes. This opacity can undermine public trust and legitimacy, threatening the integrity of elections and democratic systems as a whole. Therefore, while EMBs are encouraged to explore new AI opportunities, they must prioritize public transparency and create clear pathways for accountability and responsibility.
The AI + Elections Clinic Skills Hub is designed to foster a transparency-first approach. It serves as a platform for sharing AI-driven innovations while promoting institutional accountability by making EMBs’ AI experiences visible to both peer organizations and the general public. This is a crucial step in securing democratic legitimacy for their decision-making processes.
Rather than advocating for a one-size-fits-all approach to AI adoption, these case studies emphasize that each country can develop its own AI framework that respects local norms and contexts, all while upholding universal principles for ethical AI aligned with human rights and democratic values. As EMBs continue to devise innovative AI solutions for electoral challenges, it is essential to collect and share these insights openly, allowing others to learn from their impactful developments. This website will remain a resource for gathering and disseminating these unique experiences.
If you’re interested in exploring how EMBs navigate AI integration and develop policies for responsible AI usage, please refer to the case study series available in the AI + Elections Clinic Skills Hub.
