Documentation for transitional justice mechanisms

In looking at the effective use of civil society documentation of human rights violations for transitional justice purposes, the research revealed the following potential obstacles:

• The evidentiary requirements of transitional justice mechanisms, which may prevent human rights documentation collected by civil society actors from being admitted as evidence.

• The high volume of data that both human rights documenters and transitional justice mechanisms must manage when working in transitional justice processes.

• The serious human and data security concerns involved, which must be responsibly handled. Since technology tools can both mitigate and amplify risk, these must be deployed with care.

For transitional justice mechanisms that accept data or evidence from civil society actors, there are certain standards that must be met in order for such data to be admitted as evidence against an accused in court. Moreover, courts may be lenient in their acceptances, but the weight given to the analysis of such evidence will vary based upon how strictly it meets such aforementioned standards. To be admitted and then given sufficient weight, courts will look towards circumstances regarding informed consent of witness testimony, authenticity of data, and the chain of custody of such collected data.

This admissibility challenge manifests across the data life cycle, highlighted by the examples below:

Collection

Research participants mentioned reliability of information as sound evidence as a major concern, referencing opacity in handling data up to the point of transmission to a transitional justice mechanism, particularly in the context of ever-improving technology, through which “deepfake” videos with fabricated content are becoming more prevalent.

• To deal with these sorts of challenges, several research participants spoke to the importance of capturing metadata (see the Tools Table for some tools which do this automatically) to help to verify the authenticity of images or videos.

• Other concerns around collection as it pertains to evidentiary standards included methodologies for taking witness statements – particularly around their lack of standardisation and informed consent procedures. Several research participants were optimistic as to the ability of new technologies, and ongoing civil society training, to address these major concerns on the reliability of information and procedures for collecting witness statements.

Processing and storage

Documenting chain of custody is a critical part of meeting evidentiary standards, and while specific chain of custody requirements differ by jurisdiction, ensuring that data stands up to the scrutiny of tribunals and courts requires that documenters employ stringent processes for data storage and processing.

• Research participants raised concerns around civil society documenters’ ability to store and process data in a way that met chain of custody requirements (thus making the data more likely to be accepted as evidence). Research participants also highlighted the encryption of storage space as an important method for safeguarding the chain of custody.

Transmission

The research enquiry revealed that there is no singular, standardised method for civil society documenters to transmit information on human rights violations to transitional justice mechanisms.

• Generally, data is sent to transitional justice mechanisms via encrypted files. However, the admissibility of the data depends on the specifics of how this is done. In particular, one participant noted the preservation of metadata during transmission as crucial for the admissibility of videos and images as evidence.

The transition from traditional pen-and-paper based methods for data collection to modern technology based methods has resulted in an increased amount of raw data to be managed and analysed — which poses challenges for both human rights documenters and transitional justice mechanisms.

The challenge posed by large volumes of data manifests across the data life cycle, as highlighted by the examples below:

Processing and storage

Many research participants observed the challenge of managing data collected in large volumes of paper files.

• The time cost of digitising these files is high, so some research participants recommended using technology tools at the time of collection to eliminate the need to digitise paper files. However, others noted the need to take a survivorcentric approach that recognizes some survivors’ scepticism of technology.

• Several research participants stressed the need for transitional justice mechanisms to be able to canvass information received from a variety of civil society sources and subsequently store that information in one location. Some research participants with experience in managing human rights documentation information noted that there is no standard method or technology for canvassing and storing all available information. Some popular methods currently in use include scanning physical documents into PDF form; saving online videos, images, and audio locally; and uploading computer notes to a centralised location.

Analysis

Research participants highlighted automated tools as being useful in the face of a high volume of data.

• For example, optical character recognition (OCR) was mentioned as a tool that could enable transitional justice mechanisms to search and catalogue names, places, dates, and other relevant data points.

• Natural language processing (NLP) tools also came up in the research, but it was noted that they may be currently ill-equipped for processing data on human rights violations, given that they are at an early stage for many languages spoken by persecuted populations.

Transmission

When transmitting data, one research participant highlighted the ability to select and transmit specific parts of the dataset – rather than the whole dataset – as particularly useful in reducing the burden of a large dataset.

The final cross-cutting challenge that emerged was around ensuring the security of both data and human actors. Research participants expressed particular concern for survivors, witnesses, and documenters with videos, images, and notes of atrocities on their devices.

Research participants regarded implementing high levels of security and ensuring the anonymity of survivors, witnesses, documenters, and other actors as necessary for any technology used in a human rights documentation context.

This challenge manifests across the data life cycle, as highlighted by the examples below:

Collection

Both survivors and documenters face security risks at the data collection stage, and research participants were divided as to the safest method for collecting data.

• When discussing witness statements, those prioritising digital security considered pen and paper a safer method to limit the potential for cyberattacks and breaches of cybersecurity.

• Those more concerned with threats to human security arising from the possession of witness statements recommended documentation via a technological tool to limit exposure.

Storage

Research participants noted that both individuals storing data and individuals identifiable within the data could be at risk in the future, even if there was no immediate risk when the data was collected.

• For both evidentiary and security reasons, research participants regarded secure storage of data as fundamental to any technology for civil society documentation, and many flagged their concerns regarding the use of cloud storage.

Transmission

Most research participants highlighted that transmitting data was another moment of increased risk for documenters, victims and witnesses,, and the data itself.

• One research participant noted that some civil society documenters decline invitations to transmit evidentiary information due to human security threats associated with revealing their cooperation with a transitional justice mechanism.

• Almost all interviewees identified the ability to encrypt messaging as crucial for both chain of custody and security reasons.