Expert Roundtable on Opportunities for Artificial Intelligence and Machine Learning in Human Rights Monitoring

19 November 2024

The Geneva Human Rights Platform (GHRP) hosted the III Expert Roundtable on Digital Human Rights Tracking Tools and Databases (DHRTTDs), gathering an international group of stakeholders to address interoperability challenges and opportunities for artificial intelligence (AI) and machine learning (ML) in human rights monitoring. Co-organized with the Office of the United Nations High Commissioner for Human Rights (OHCHR) and HURIDOCS, and as part of an ongoing initiative led by GHRP, this event builds on last year’s Expert Roundtable, which underscored the growing significance of digital human rights tools and databases in achieving efficient, coherent monitoring systems.

This year's roundtable, held at Villa Moynier in Geneva and online, welcomed over 30 experts, including DHRTTD developers, AI specialists, policymakers, and human rights practitioners. Participants explored the theme, ‘Interoperability through Artificial Intelligence (AI) and Machine Learning (ML)’, with sessions that examined both the technical and ethical dimensions of integrating AI into human rights tracking.

‘Interoperability is the backbone of a cohesive human rights monitoring system,’ noted Felix Kirchmeier, GHRP’s Executive Director. ‘Achieving it means reducing redundancy, enhancing real-time data sharing, and fostering a more unified approach to monitoring globally.’

The roundtable examined frameworks for standardized data formats, common protocols, and sustainable funding models to enhance information-sharing across diverse tools while preserving each tool’s unique functionalities. These discussions aimed to reduce fragmentation and enable more comprehensive and efficient human rights documentation and tracking.

‘Interoperability is key to connecting the UN system with regional and national human rights processes,’ underlined Andrea Ori, OHCHR, Chief of the Human Rights Treaties Branch, a.i. ‘By making digital tools practical and responsive, we reduce fragmentation in data tracking, enhance engagement with human rights mechanisms, and strengthen national capacity for effective advocacy and implementation of international human rights obligations.’

A significant focus was also placed on the dual role of AI and ML in advancing and challenging the field of human rights monitoring. AI and ML offer significant capabilities for streamlining data processing, automating repetitive tasks, and generating predictive insights that can help anticipate and address human rights violations before they escalate. Participants explored AI’s functionality for clustering, tagging, and pattern recognition within large datasets, which enables a more responsive and real-time approach to tracking and responding to human rights issues.

‘The human rights community must be mindful of both the promises and perils of AI’, highlighted Yolanda Booyzen, Director of Development and Communications at HURIDOCS, ‘By keeping the human in the loop, and through building ethical human rights-centred tools, we can increase the efficiency and therefore the impact of the human rights ecosystem’.

As such, the roundtable also tackled the ethical and practical challenges associated with integrating AI and ML in human rights tracking. Participants explored critical issues such as algorithmic bias, which risks perpetuating existing prejudices if not carefully managed, and transparency, noting the challenges of ensuring that AI-driven decisions are understandable and accountable. Privacy and data security were emphasized as paramount, especially given the sensitive nature of human rights data. This focus on ethical AI aimed to balance the technology’s potential with the responsibility to safeguard human rights values.

‘The potential of AI to automate and analyze vast amounts of data can be transformative for human rights monitoring,’ said Domenico Zipoli, GHRP Project Coordinator. ‘But we must ensure it serves as an ethical and trustworthy partner in this work, not a source of bias or opacity.’

In the final session, participants discussed interdisciplinary collaboration as a pathway to innovation and responsible AI use in human rights work. The roundtable concluded with actionable next steps, setting a clear path for future initiatives aimed at advancing digital human rights monitoring. Key action points included:

1. Expanding the Directory of DHRTTDs: GHRP committed to further populate its DHRTTD Directory to capture a more comprehensive range of tools in this space.
2. Mapping Interoperability Potential: A mapping exercise will be conducted to identify potential cooperation and interoperability among different human rights tracking tools, databases, and measurement projects, distinguishing specific needs and functions to improve interoperability.
3. Establishing a Dialogue Space: Plans are underway to create a collaborative forum for AI, human rights, and SDG experts to engage in directed, ongoing discussions, potentially within the AI for Good Summit framework.
4. Forging Knowledge Exchange Partnerships with Universities: GHRP aims to establish informal partnerships with institutions like ETH Zurich and the University of Zurich to facilitate knowledge exchange on AI applications in human rights.
5. Collecting and Grouping AI/ML Use Ideas: Participants will contribute ideas on potential uses of AI and ML in human rights and SDG tracking, which will be grouped by functionality to identify areas suitable for automation.
6. Compiling a List of Beneficial AI Functionalities: Based on the collected ideas, GHRP will compile a list of specific AI functionalities that could enhance human rights and SDG monitoring.

The event also highlighted the upcoming Working Paper AI Decoded: The Foundational Concepts and Applications for Human Rights and SDG monitoring, a practical guide to AI and ML terms and concepts designed to facilitate cross-sector collaboration on AI in human rights. ‘Our goal with AI Decoded is to bridge the knowledge gap,’ explained Milica Mirkovic, Project Assistant at GHRP. ‘We want practitioners to feel confident in engaging with AI concepts, ensuring that technology serves human rights rather than complicating it.’

The outcome document, to be published by GHRP, will summarize these discussions and provide guidance on responsible AI use, supporting global efforts to build a transparent, effective human rights monitoring framework.