29 Nov ArcGIS & Hearsay: Challenges of Incorporating New Technologies into Existing Evidentiary Frameworks
[Chris Carpenter is a lawyer and researcher in international law. She holds a J.D. from the University of Pennsylvania Law School and a master’s in International Relations and Politics from the University of Cambridge.]
The fundamental goal of a fair trial—at least in theory—is the discovery of truth, and the finding of a verdict on the basis of that truth. But in an age where information flows freely and ubiquitously across the globe, sieving truth from the mass of information readily available is somewhat like drinking from a firehose. In the courtroom, this creates unique challenges as conventional rules of admissibility are confronted with the proposed inclusion of new technologies, which are used for both investigative purposes and for demonstrative purposes at trial.
So-called ‘open-source information’, this publicly available content has been incorporated into criminal trials with increasing popularity. One example is through the inclusion of ArcGIS mapping technology into both the collection and presentation of evidence. ArcGIS is an online, web-based geographic information system that contains spatial data in the form of web-maps, web-applications, or maps. A final product made with ArcGIS shows the map in ‘layers’, with each layer containing spatial and physical attribute information for a given location. This content can take the form of imaging, text, data models, or even video. The use of this technology in investigating human rights violations is not new (see here, here, and here, for examples). However, explicit consideration by international tribunals, like the International Criminal Court (ICC), as to how including these sources implicates various evidentiary concerns is less robust—although these technologies no doubt bring with them a myriad of challenges pertaining to the rules of evidence.
In addition to creating authentication and relevancy issues for courts to contend with, a particularly complicated evidentiary rule to adapt to the inclusion of this information at trial is hearsay. The term ‘hearsay’ refers to an out-of-court statement offered at trial to prove the truth of the matter asserted. Such statements are generally inadmissible—barring a series of convoluted exceptions that vary by judicial system. This determination is rooted in the policy interest of preserving truth at trial, as out-of-court statements do not allow the opposing party to confront the speaker (called the ‘declarant’), and as such, cannot be properly evaluated for their truth content.
Hearsay rules vary between judicial systems, which weigh the pros and cons regarding admitting out-of-court statements differently. In international criminal law generally, judges possess broader discretion to admit evidence than is seen in domestic courts, which corresponds to the enhanced challenges of investigating international crimes. Under the Rome Statute, the International Criminal Court has no explicit hearsay rules. Rule 69(4) of the Rules of Procedure and Evidence instructs the Court to consider ‘any prejudice that [the evidence] may cause to a fair trial’ in its admissibility decisions, which is comparable to the requirement in Rule 403 of the US Federal Rules of Evidence (FRE) that evidence be more probative than prejudicial. However, emphasis is placed on the ability of the court to assess evidence ‘freely’, per Rule 63(2), and the Court makes admissibility determinations at the end of the trial process. Thus, there is no ex ante challenge for out-of-court statements being used for their truth content at the ICC. At the other end of the spectrum, common law states like the U.S. rely on an extensive web of rules dictating hearsay’s admissibility—that is challenged in advance of admission. In the U.S., for example, hearsay is generally inadmissible under FRE Rule 802, save for a number of designated hearsay exceptions and non-hearsay exclusions under Rules 803 and 804.
While courts approach hearsay’s admissibility very differently, the root concern is the same: the appropriate weight to be given to a piece of information that is gained from a source other than the initial speaker or author. While the ICC may default to admitting a wider array of evidence—including hearsay—judges may then adjust the probative weight given to these pieces of evidence. A judge sitting in a U.S. federal court may be less likely to admit the evidence upfront, but if the piece of hearsay is found to fit within a hearsay exception, it is often admitted without the same constraints. The point is not that one of these approaches is better or worse than the other, rather that both have struggled in their own way with hearsay derived from technological advances that have progressed more quickly than the evidentiary rules under which they are assessed.
Regarding challenges posed by technological advances specific to hearsay, a threshold challenge is determining what even constitutes a ‘statement’. Often, the hearsay statement in question is verbal. But the term hearsay also applies to emails, text or chat messages, and other forms of digitally communicated writing, as well as sound captured on a video or audio recording. Regardless of its form, however, the ‘statement’ requires a declarant to fall under the hearsay rules. Historically, courts have refused to consider a machine capable of being a ‘declarant’ for hearsay purposes. But this does not mean machine-based statements are immune to hearsay concerns, especially in a world where machine-statements and human-statements are often infused into a shared final product – like an ArcGIS map. In this context, do the authors of the various, publicly available layers of data ‘speak’ through their creations? It is here we see courts struggle to adapt long-standing principles to new modes of conveying evidence.
Hearsay Concerns & Mapping Tools Before the ICC
Open-source information has been included in international criminal trials from their inception, with the introduction of video evidence documenting Nazi concentration camps being shown during the Nuremberg Trials. The ICTY cases of Prosecutor v. Kristić and The Prosecutor v. Mladić, the OTP incorporated satellite imagery, in addition to photo and video evidence, in its case before the Trial Chamber. In recent decades, courtrooms at the ICC have already hosted numerous technological advancements. For example, in Prosecutor v. Ntaganda, Prosecutor v. Abu Garda, and Prosecutor v. Ruto and Sang, satellite imagery was used to aid in identifying locations of alleged war crimes—and to demonstrate the extent of the damage. However, this evidence was not challenged on hearsay grounds, nor would such challenges have been likely to succeed. As is true in U.S. courts (see below), satellite imagery is largely treated like a photograph for hearsay purposes. It makes no assertion, it has no speaker, and therefore, its admissibility is not challenged as hearsay.
However, as geospatial technology progresses to include not only photograph-like imagery, but also audio, video, data modeling, or a combination thereof, hearsay concerns become more applicable. A recent and primary example is the case of The Prosecutor v. Al Mahdi, which concerned the destruction of cultural heritage sites in Mali. At trial, the prosecution relied in part on open-source videos, photographs, and satellite images from YouTube, Google Earth, and other internet sources, which had been compiled into a digital platform to create side-by-side comparisons of the sites before and after they were destroyed. The multi-faceted, geographic imaging that resulted is very similar to how ArcGIS can be used to visually represent evidence in a courtroom. However, as Al Mahdi pleaded guilty, proceedings concluded before evidentiary challenges to this platform were made—due to the ICC’s ex post process for admission of evidence.
Virtually the same form of multi-faceted geographic imaging was used in the case of The Prosecutor v. Al Hassan, but helpfully, as of June 2021, we have a decision on the admissibility of specifically open source exhibits into evidence. While the Trial Chamber notes in the decision that it is not ‘issuing a ruling’ on admissibility, rather a mere ‘recognition’ of the admission, this still provides the opportunity to see what objections to this content arise. Among over half a dozen objections to the open-source materials is the defence counsel’s objection that ‘submission of certain items would be prejudicial “due to non-attribution of the source of the information (e.g., anonymous hearsay)”’. As the trial is still ongoing, and the ICC rules on admissibility objections at the conclusion of the trial, we will have to wait and see how this hearsay objection resolves—namely if it meets a different fate than its American counterparts.
A Higher Bar for Hearsay Does Not Mean Lower Chance of Admissibility
In spite of its more robust array of rules surrounding hearsay’s admissibility, the more formalized common law approach has fared no better in addressing hearsay concerns emanating from this technology. If anything, the kinds of technology examined for hearsay issues in U.S. courts are more distant from ArcGIS than those used before the ICC.
The body of case law considering hearsay’s applicability to machine-made statements more broadly, which would presumably be applied to the admissibility of ArcGIS when it is challenged on hearsay grounds, is known as the ‘machine-generated testimony doctrine’. Under this doctrine, the closest proxy for ArcGIS software considered by any U.S. Circuit is United States v. Lizarraga-Tirado, where the Ninth Circuit considered whether the defendant was North of the U.S.-Mexico border at the time of his arrest. At trial, the prosecution used both a Google Earth image of the area where the defendant was arrested, as well as Google’s GPS function to place a ‘tack’ on the map indicating the coordinates at which the defendant was arrested. The defence objected to both as inadmissible hearsay. In reviewing the trial court’s decision to admit the Google Earth map, the Ninth Circuit likened the satellite image to its hearsay case law regarding photographs and held that a satellite image likewise ‘makes no assertion’, and thus concluded that satellite images cannot be inadmissible hearsay. However, the court did note that the ‘tack’ on Google Earth image could have been considered hearsay, if it had been ‘placed manually and then labeled’. As the tact was in fact ‘automatically generated by the Google Earth program’, hearsay was not ultimately found in this case—but the court did acknowledge the labeling process to be a more challenging issue.
Yet, when a closer call arose in Walters v. State, the Court of Appeals of Nevada also declined to find hearsay where an exhibit of a Google Earth image involved a ‘tack’ that, in this case, included the label ‘Douglas Walters Home’. Citing to Lizarraga-Tirado, the court reasoned it could not say whether the tack was ‘generated automatically by a computer or was placed on the image manually’.
Unique Evidentiary Challenges Posed by ArcGIS: Concerns and Recommendations
While the ICC has begun to rely on more advanced digital imaging to present evidence at trial, we are still a long way from having sufficient material to truly assess what challenges new technologies like ArcGIS pose for existing evidentiary landscapes. The scope of concerns for hearsay purposes implicated by this software is enormous. Were these platforms to be used in a courtroom to their full potential—complete with audio, video, and written content, in addition to the geographic imaging—the implications strike to the very core of hearsay’s concerns. The data-layer files that make ArcGIS so unique also make images rendered with this technology inherently problematic from a hearsay perspective. The layers of information included in any given map may come from an untellable number and variety of sources—of differing degrees of accuracy. The layers of open-source information are gathered into a final product that is then placed in front of a jury and purported to convey objective fact, currently without any requirement for authentication or means of confrontation (assuming courts treat ArcGIS imaging consistently with how they have treated the other new forms of technology examined here).
However, the ultimate challenge is how to use existing rules of evidence to confront these concerns in relation to technology like ArcGIS software. As the ICC’s consideration of hearsay essentially collapses into a probative-versus-prejudicial balancing test, this form of evidence may well still be admissible in spite of its augmented hearsay concerns. While the audio and video content contained in ArcGIS documentary evidence is laden with hearsay, which by itself would indicate a lesser degree of reliability, and as such a diminished probative value, the probative side of the scale is substantial. This goes to the core rationale for the lower admissibility thresholds among international criminal tribunals. International crimes are exceedingly challenging to investigate—they cross borders, span multiple years, involve tens of thousands of people, and are often hard to document in real time. Meanwhile, the prosecutor must demonstrate the perpetrator’s guilt beyond a reasonable doubt—the highest burden of proof available. This makes the kind of information available in an ArcGIS map of immense probative value in a field where reliable probative evidence is often a scarcity. Thus, barring other concerns of reliability and authenticity, hearsay is an unlikely barrier to admission. However, this does not mean a hearsay inquiry surrounding this kind of evidence is not important to preserving the trial’s legitimacy.
To this end, a promising new development is underway. As the Al Hassan trial comes to a close, the ICC holds the opportunity to impart one of the first definitive rulings on the admissibility of ArcGIS-adjacent evidence—and in the wake of a hearsay objection, no less. A step toward clarity may soon be coming, but this is a far cry from resolution. If an examination of ArcGIS in the courtroom can show us one thing, it is that the world of digital evidence confounds many of our existing rules and doctrines. They must adapt to meet this growing technologization of the courtroom—and the law more generally—if the vital evidentiary traits of a fair trial are to be preserved.