In 2014, while working with the organization Act for the Disappeared in Lebanon, a researcher encountered a mother whose son had vanished during the country’s civil war. Decades after the conflict ended in 1990, she kept his room exactly as he had left it, with his clothes washed and neatly folded, waiting for a return that never came. This encounter, and the broader struggle of the 17,000 people estimated to have gone missing in Lebanon, serves as a poignant reminder of the human cost behind historical statistics.
Years later, this experience prompted an exploration into whether artificial intelligence (AI) could assist in the search for truth and justice. While technology cannot replace the political will required to resolve long-standing disappearances, it is increasingly being tested as a tool to navigate the vast, complex archives of human rights violations.
Technological Applications in Human Rights Investigations
Over the past three years, researchers have begun integrating AI into transitional justice efforts to manage large datasets that would otherwise take years to process. These applications vary significantly depending on the specific needs of the investigation:
- Identifying Clandestine Graves: In Mexico, researchers have utilized machine learning combined with satellite imagery to pinpoint potential burial sites.
- Archival Analysis: The Cruzar project in Uruguay employs AI to assist in the examination of dictatorship-era archives, working in collaboration with the families of the disappeared.
- Data Processing: AI tools can convert scanned historical documents into searchable text, helping investigators identify names, locations, and potential connections between disparate records.
The Risks of Algorithmic Investigation
Despite these advancements, AI is not a panacea. The technology is prone to errors, such as misinterpreting context or producing plausible but incorrect conclusions. Furthermore, the "black box" nature of some systems makes it difficult for investigators to understand how a specific result was reached, necessitating rigorous human verification of all AI-generated leads.
Establishing Accountability and Safeguards
The integration of AI into justice efforts raises critical questions about responsibility. If an AI tool fails to link records due to transliteration errors, or if it is used in ways that expose survivors to further harm, the lines of accountability remain blurred. Moreover, as AI is increasingly used in modern surveillance and military targeting, the field of transitional justice must also grapple with how to reconstruct decisions made by algorithms that may have contributed to new violations.
In July 2026, UN Special Rapporteur Bernard Duhaime addressed these challenges, calling for stronger regulation of technology companies and the implementation of mandatory human rights impact assessments. Experts emphasize that the path forward requires a collaborative approach involving technologists, legal professionals, and, most importantly, the affected communities themselves.
As the field evolves, organizations are encouraged to ask fundamental questions before deploying new tools: Is the use of AI truly necessary? Have the original contributors of the data consented to this type of analysis? And what safeguards are in place to ensure that the pursuit of truth does not inadvertently cause further suffering to those still waiting for answers.
