By Dean P. Davison, M.B.A., LL.B., CFE
The Canadian experience shows that many issues related to the admissibility of digital images in court proceedings remain unresolved.
Still photos and video images have always been compelling when presented as evidence in court. For that very reason, courts have set high standards to ensure the integrity of those images. The advent of digital images raises additional concerns, because the images can so easily be manipulated.
In Canadian courts, images as evidence were originally admitted to assist and complement a witness’s testimony. However, a 1996 decision by the Supreme Court of Canada (R. v. Nikolovski) allowed video evidence to speak for itself as a “silent witness” even if an eyewitness gives conflicting testimony.
Admission of image evidence is linked to a common law test (R. v. Creemer and Cormier) developed in 1968—obviously well before the digital age. This test assesses the accuracy of image-based evidence in representing the facts, and assesses its genuineness and authenticity.
A good example of the application of the test by the courts is the case of R. v. Mahoney (1976). In Mahoney, the accused was a hockey player on trial for assaulting another player during a game. The presiding judge said that because the video evidence was played back in slow motion, it did not accurately portray the facts at issue. He felt that the slowing of the assault made it look too deliberate and particularly violent.
This test identifies three groups of individuals that can swear that an image is authentic: a witness to the event, a cameraperson who took the image, and an expert on the camera system.
In light of the ease of editing digital images and the inherent problems with eyewitness testimony, however, depending on the first two groups (essentially both witnesses) is problematic, as they are unlikely to notice altered or unauthentic images when testifying perhaps years later. Thus, an expert may be the best way to ensure authenticity.
But in digital images, the editing is undetectable and strikingly simple, raising the question of whether even an expert can tell when digital images have been manipulated. Accordingly, there is a growing need for valid, dependable, and consistent authentication procedures for digital image evidence.
As such, authentication of digital images in criminal or civil cases without a camera operator or an eyewitness to corroborate should follow the Canada Evidence Act (CEA), as this is the best option open to the courts for authentication.
This means that a person responsible for the computer system in question must give sworn testimony that the computer system was at all times operating properly or that any operating problems was not such as to affect the accuracy of the image in question. But proper system functioning alone is not sufficient to prove image authenticity. It must be shown that there was no opportunity for alterations.
In a practical situation, digital video evidence that was captured with a standard computer compression program and turned into an .AVI or Mp3 file for storage would be suspect if the opposition challenged the system’s integrity. Continuous integrity is achieved if there is restricted access to the recording system and if transfer of the data is done in such a way that the operator can swear to its authenticity.
As video-editing software is readily available and the result is difficult to detect, nothing but secure and continuous protection—a chain-of-custody assurance—would be sufficient to ensure the integrity of the evidence. This integrity test should have to be done regardless of witness corroboration, and the burden should lie with the party introducing the evidence. Currently, for many digital image-capturing systems, this is not a reality as multiple users are involved who interact with most digital image systems in the normal course of business.
It is also worth noting that in most criminal cases, it will be either the police or an interested party creating the lion’s share of digital images used by the government as evidence. For example, the store clerk in Nikolovski had control over the video that was used as evidence. In this particular situation, the store used traditional videotape and authentication was not such an issue, but, if it were, what would prevent the shop owner or clerk from altering the image with editing software?
If anything, an involved party’s evidence should be more suspect than the evidence the police have, as the police have no personal interest or desire for revenge or retribution. Additionally, because digital images come in a variety of different formats, it could be assumed that it would be much easier for the owner of the equipment to alter the digital images as compared to the police, who may not be familiar with the particular digital equipment or format.
Furthermore, the police have special divisions that deal with technology and digital images. Thus, the officer who seizes the tape would have had to ask the technical officer to alter the evidence for him. Although alterations may be highly unlikely and many proprietors would not and could not alter digital images, alterations are a possibility and have to be contemplated by the court.
The CEA does allow industry standards to be used as evidence that the electronic document was not altered. For example, with analog video, it is standard practice to initial, date, and sign the videotape and store it in a secure area until it is required as evidence. This usually ensures that the video evidence is admissible at least with regards to authenticity.
However, a major concern with digital image evidence in Canada is that there are no industry standards as to recording, storing, and retrieving. Without industry standards, how is an expert going to prove the authenticity of the digital images? Some methods of ensuring authenticity, such as watermarking, are starting to appear, but they are proprietary and, thus, nonstandard.
Moreover, the process of transferring the evidence from the computer system that captured the images to a format suitable for the courts is currently without standards. Many digital images are now transmitted via intranets or networks with an operator working from a personal computer to capture and control the images.
This creates a serious concern for the integrity of the image and its storage, notwithstanding the very real potential of hackers. Although a higher and consistent standard creates a more onerous burden for the law enforcement community, it seems appropriate in light of the Supreme Court of Canada determination that the first step in admissibility is to ensure that the evidence has not been tampered with or altered.
Currently, converted videotape or hard copies of a converted digital image are being offered as evidence and the original digital image is not being challenged. However, if opposing counsel start to ask the right questions, there may be some real challenges for the courts to accept digital image evidence as authentic. If digital images become inadmissible because their authenticity cannot be proven, that may push the industry to create universal systems and protocols to ensure that digital image systems are dependable. If the courts require a standard for authentication of digital images as evidence, manufacturers will have to comply to continue to sell their products.
But waiting for the courts or the legislature to establish a standard is risky. The security industry would do well to create a standard that the courts understand and can depend on, rather than waiting until courts impose their own
Dean P. Davison, M.B.A., LL.B., CFE (Certified Fraud Examiner), is a lawyer with Watson Goepel Maledy LLP, in Vancouver, British Columbia. He is a member of the Canadian Pacific Chapter of ASIS International.
@ For information on how this issue of digital image admissibility is evolving in U.S. courts, go to www.securitymanagement.com and link to an article by Rebecca Levy-Sachs and Melissa Sullivan of Robinson & Cole.