The loss of life is a tragic reality in armed conflicts around the world, leaving thousands of unidentified victims in varying states of decomposition. Some bodies are mutilated, desecrated, or abandoned without funeral rites, while others are buried in mass graves with no record of their identity. International humanitarian law establishes obligations for warring parties to collect and search for the dead, prevent the desecration of remains, ensure dignified burials in accordance with religious and cultural traditions, and account for those who have died.

Technological advancements are opening new possibilities in forensic investigations, particularly in the use of artificial intelligence and machine learning to assist with identifying victims of war. Although these technologies have been applied in other areas of forensic science, their role in humanitarian forensic action remains largely unexplored. In recent years, forensic experts and technology specialists have developed AI-driven tools such as Skeleton ID, Commingled Remains Analytics, and forensic facial imaging software to improve the identification of missing persons and war victims.

Artificial Intelligence in Humanitarian Forensic Action

Artificial intelligence and machine learning have already been integrated into disaster response efforts, aiding in risk assessment, hazard detection, and the identification of vulnerabilities. During the COVID-19 pandemic, AI-based imaging tools helped detect pneumonia lesions with an accuracy rate of 83 percent. AI and machine learning are also used in conflict settings for geospatial intelligence, open source material analysis, and the prediction and identification of mass graves in post conflict areas.

When applied to forensic identification, AI has the potential to improve the accuracy and speed of matching remains with existing records. Traditional methods such as visual recognition, fingerprint matching, DNA testing, and dental comparisons often rely on preserved materials for comparison. In cases where only skeletal remains are available, these methods may not be effective. AI driven forensic tools can assist forensic practitioners with image recognition, pathology classification, and the enhancement of degraded images to aid in identification.

The Korean War Identification Project

Following the 1953 Korean Armistice Agreement, efforts were made to recover and exchange the remains of soldiers killed in the war. However, many US soldiers were not repatriated before the closure of the North and South Korean border, leading to commingled remains that made recovery and identification difficult. To address this challenge, the United States developed predictive models and full spectrum identification algorithms that analyse isotopic values in tissues. These models enable forensic experts to separate and identify remains, even when they have been commingled. The use of software programs such as Commingled Remains Analytics has also helped categorise and distinguish remains into distinct individuals.

Advancements in Forensic Facial Imaging

Machine learning has played a significant role in the development of forensic facial imaging tools, which assist with facial approximation and photographic superimposition. Research groups such as Face Lab at Liverpool John Moores University have created AI driven facial recognition technologies that generate three dimensional craniofacial representations and databases of anatomical traits. These tools help forensic scientists reconstruct facial features from skeletal remains, improving the likelihood of identification.

Skeleton ID and Skeletal Analysis

Skeleton ID is another AI based tool that uses physical anthropological methods such as biological profiling and comparative radiography for identification. However, AI driven skeletal identification techniques have limitations, particularly in cases where remains have been severely fragmented due to bombings or environmental degradation. A multidisciplinary approach that incorporates multiple lines of forensic evidence, including dental, genetic, medical, and geographical data, is often necessary to achieve reliable identification.

Legal Obligations Under International Humanitarian Law

The Geneva Conventions and other international humanitarian law treaties establish clear obligations for warring parties to preserve the dignity of the deceased and their families. These include:

– Search and Recovery of the Dead: The Geneva Conventions require warring parties to take all possible measures to locate, collect, and evacuate the dead. Agreements should be made to allow search teams and humanitarian organisations to assist with recovery efforts.

– Dignified Burial and Grave Protection: Proper medical examinations must be conducted before burial to establish identity. Graves should be marked and maintained permanently, and burials should align with the religious practices of the deceased. Individual burials are preferred over mass graves unless circumstances require otherwise.

– Humane Treatment of the Dead: International law prohibits the mutilation or desecration of remains. The display of graphic images or videos of war victims on social media can also violate humane treatment principles, as exposure to public curiosity may cause additional distress to families.

– Return of Human Remains: While the Geneva Conventions do not establish a universal obligation to repatriate remains, they encourage warring parties to facilitate the return of bodies to their families when possible. Several United Nations resolutions have reinforced this principle.

Opportunities and Challenges in AI Based Forensic Action

Artificial intelligence presents both opportunities and challenges in humanitarian forensic work.

– Enhanced Accuracy and Speed: AI can automate repetitive tasks such as image recognition and data analysis, reducing forensic backlogs and accelerating identification processes. This can minimise delays in returning remains to families and allow them to conduct proper burials in accordance with their cultural and religious beliefs.

– Bias and Ethical Concerns: AI driven forensic tools can reinforce existing biases if they are trained on limited datasets. Discriminatory algorithms could result in misidentifications, causing distress to families. Developers must ensure that AI models are designed with diverse and representative datasets to prevent such errors.

– Data Privacy and Security: The collection and storage of genetic and biometric data raise ethical concerns, particularly in conflict zones. Families of the deceased should provide informed consent before their relatives’ data is used in forensic investigations. Additionally, safeguards must be in place to prevent data theft, misuse, or unauthorised access.

– Limitations in Conflict Zones: The use of AI in war zones is often hindered by logistical constraints. Non state armed groups may lack the resources or willingness to employ forensic AI technologies, leading to inconsistencies in how forensic obligations are fulfilled across different conflicts.

Policy Recommendations

To maximise the benefits of AI in humanitarian forensic action, governments and international organisations should implement the following measures:

– Develop International Standards for AI in Forensics: A legally binding global framework should be established to regulate the ethical use of AI in forensic identification.

– Create Ethical Guidelines for AI in Humanitarian Contexts: International humanitarian organisations should develop policies that align with human rights principles and ensure that AI technologies do not exacerbate trauma or distress.

– Provide Training for Forensic Experts and AI Developers: Training programs should educate forensic practitioners on AI applications in forensic science, while also ensuring that AI developers understand the ethical and legal implications of their technologies.

Conclusion

Artificial intelligence has the potential to revolutionise forensic identification and ensure dignified treatment of the dead in armed conflicts. By accelerating forensic investigations and improving accuracy, AI can provide closure to families and uphold the principles of international humanitarian law. However, these technologies must be implemented with caution, ensuring that ethical considerations and human oversight remain central to their use. A human centred approach that respects cultural traditions and prioritises dignity in death is essential in advancing the role of AI in humanitarian forensic action.

Republished courtesy of Lieber Institute of West Point

Note content has been edited