My colleagues and predecessors here at the School of Government have written about video evidence many times over the years, summarizing the basic rules and significant cases in posts available here, here, here, here, and here.
Recently, though, I’ve been getting questions about a relatively new but increasingly common type of video evidence: high-tech, app-controlled, and remotely stored videos taken by automated devices ranging from doorbell cameras to wifi-enabled, cloud-connected, teddy bear spy cams. Do the old rules still work the same way for these new video tools? Is it substantive or illustrative evidence? If it’s substantive, how is it authenticated? Is a lay witness qualified to testify about how these cameras work? Does the proponent need the original video? Come to think of it, what is the “original” of a video that exists only as bits of data floating somewhere in the cloud…?
The Good Old Days
Once upon a time, there were only a few kinds of video that ever saw much use in criminal trials — dash cam videos and recorded interviews verified by the officer who observed the event, or commercial surveillance videos authenticated by a knowledgeable store employee. Laying the foundation for one of those videos was pretty straightforward. The officer or employee would pull the disc (or “tape” – go ask your parents) out of the camera, put a sticker on it, make a copy for discovery, place the original into an evidence locker, and then bring it to court for trial.
In this post from 2016, Jeff Welty observed that there has been an explosion in the amount of video being created and shared in recent years, but the existing rules and case law for authenticating and admitting those videos still seemed to be holding up pretty well. Even when it’s a video taken by someone using a digital device like a cell phone, the “original” video is usually still right there on the device so it can be easily copied, verified, and authenticated.
But now we have videos that are: (i) captured automatically; (ii) by an unattended doorbell or teddy bear; (iii) stored on a cloud server halfway across the world; (iv) accessed remotely and after-the-fact; (v) through a cell phone app; (vi) by a victim or other witness who has absolutely no idea how any of the underlying technology works.
That just feels… different.
Objection: Authentication, Foundation, and Everything Else
I participated in a prosecutor training program earlier this week where several attendees confirmed that they were beginning to see this kind of video more often, and a couple of recent appellate cases seem to bear that out. See, e.g., State v. Vetter, 257 N.C. App. 915 (2018) (state’s evidence included video captured by victim’s home security system and observed through a phone app). However, the admissibility of the video was not one of the issues addressed in Vetter, and I haven’t come across any other North Carolina cases that provide a detailed discussion about the foundation requirements for it.
Until we get that case, an unpublished decision from Texas offers a helpful example of how this might play out at trial and on appeal. See Chatman v. State, 2018 WL 6629531 (Tx. App. 2018) (unpublished) (used here for illustrative purposes only, of course).
In Chatman, the victim was at home when he got a notification on his cell phone through the Ring doorbell app. The victim initially answered the Ring through his phone, and then came to the door and opened it, at which point he was violently assaulted. Portions of the assault and its aftermath were captured in three video clips by the Ring camera. The victim later used the app on his phone to access his personal account in the cloud, downloaded the videos of the assault, and emailed them to law enforcement. At trial, when the state offered the videos into evidence, the defense made a lengthy objection covering everything from the rule of completeness to being unfairly prejudicial and needlessly cumulative, but the core authentication issues being raised were these:
[T]his video footage comes from ring.com, or it comes from a dot com of some sort. And in reviewing the discovery, what happened is the [complaining witness] said, Here you are, law enforcement, and was able to log into a cloud and then retrieved some videos and then e-mailed them to law enforcement […] and I don’t believe that law enforcement ever had access to the actual account to actually download the information themselves. […] [I]t’s my understanding that there was no system that it was ever recorded to. Let alone, does this witness have the ability to talk about how this even works? That, I believe that this would be beyond his expertise. […] [I]t would limit our cross-examination as it relates to maybe making some challenges about this device and whether or not it should be considered a credible device because we don’t have someone who can speak intelligently about the ins and outs of the actual mechanism. So we have an issue as relates to my client’s constitutional right to cross-examine because we will not fully be able to cross-examine this witness about the mechanism. We don’t think that he is the proper authority to actually authenticate it.
That objection does a pretty good job of articulating the broad questions and concerns listed above and applying them to the case at hand. Nevertheless, the trial judge in Chatman overruled the objection and admitted the video, and the appellate court affirmed that decision:
The events occurring outside the front door were captured on Greer’s home video surveillance system. The camera at the front door was motion activated. Video images captured when the camera was activated were stored in a password-protected “personal file” in the “cloud.” Greer could view the videos by accessing an “app” on his mobile phone. […] Greer described the events depicted in the video clips and testified the camera at his front door captured three clips of the offense that occurred outside the apartment, he recognized the people and voices on the video clips, the video clips accurately depicted the events occurring outside the apartment on the day of the offense, and he believed viewing the video clips would help the jury understand his testimony. In this case, Greer was a witness with “personal knowledge who observed the scene,” and the State properly authenticated the video footage through Greer’s testimony.
If that case were being decided in North Carolina, would it turn out the same way? I think it would, but let’s look at a few issues in more detail.
Illustrative Purposes? Easy Answer
The Chatman decision never used the phrase “for illustrative purposes,” but that seems to be the court’s primary basis for admitting the video. The witness was physically present when it happened, the video accurately depicted what he experienced, and it will help the jury understand his testimony. As covered in the other blog posts linked above, when a video is being offered for the limited purpose of illustrating the witness’s testimony, the only foundation required is that the witness be able to say that it is a fair and accurate depiction of the event, and it would be helpful in explaining his or her testimony to the jury. It doesn’t really matter whether the video was taken by an automatic Ring doorbell linked to the cloud or a tamper-proof analog camcorder locked up in a vault. As long as the witness testifies that it’s fair, accurate, and helpful, the foundation for illustrative purposes has been laid.
Substantive Purposes? More Complicated
At least one of the videos in Chatman contained some events that were not personally observed by the witness, since they happened out on the porch before he opened the door. For argument’s sake, let’s assume the state intended to offer that video as substantive evidence to establish a key detail in the case that no other witness testimony or evidence could prove. Would that be allowed? Admitting a video as substantive evidence requires a different type of foundation, and raises different concerns.
1. Camera Functioning Properly
The first requirement is that the witness must be able to say that the camera was functioning properly at the time the video was recorded. Importantly, however, being able to testify that the camera was working does not mean the witness must also be able to explain how it works. As long as the witness knows that the system does, in fact, function correctly (Rule 901(b)(1), “Testimony of Witness with Knowledge”) and that it reliably and accurately records whatever happens in front of the lens (Rule 901(b)(9), “Process or System” that produces an “accurate result”), the first foundation element should be satisfied. Applying the same rules and logic, prior cases have consistently allowed a lay witness such as a store employee to authenticate videos recorded by complex, automated surveillance systems installed in businesses and restaurants. See, e.g., State v. Snead, 368 N.C. 811 (2016); State v. Fleming, 247 N.C. App. 812 (2016); State v. Ross, 249 N.C. App. 672 (2016).
How would the proponent satisfy this part of the foundation for a doorbell or teddy bear camera at the lay witness’s home? It could be as easy as asking the witness: “was the camera working properly that day? How do you know?” For example, has the witness previously seen herself coming and going on it? Heard herself talking to someone? Was it always an accurate recording on those prior occasions? Ever see a package get delivered when she wasn’t home? Was it waiting there on the porch when she got back? Were the date and time stamps accurate? Does she have any reason to think that it was not working the same way on the day of the crime? The answers to those questions offer a common sense basis for showing how and why the witness can be confident that the camera was functioning properly on this occasion as well.
2. Same Video, Not Altered
To meet the second foundational requirement, the proponent must establish that the video being offered in court is the same one recorded by the camera, and the video has not been altered. In the old days, when a disc or tape could be ejected from a camera and safely stored until trial, that was easy to do. But with digital video saved on a remote server and then downloaded through an app, how can we be sure?
Based on the “automatic recording” case law cited above, it should be sufficient for the witness to testify that he or she accessed the stored video through her account, downloaded or saved it in the normal way, and the video here in court is the same one that he or she retrieved. See, e.g., Snead, 368 N.C. at 816 (authentication requirement satisfied by witness’s testimony that video in court was “exactly the same video that he saw on the digital recorder”).
Is it possible that before the witness performed that download, someone might have hacked into the Ring server and replaced the real video with a perfect “deep fake” intended to frame an innocent person? Well, sure, anything’s possible, but unless there is a genuine reason to think that it did occur, the witness’s testimony vouching for the video ought to be sufficient. See Id. at 815 (“Because defendant made no argument that the video had been altered, the State was not required to offer further evidence of chain of custody”); Ross, 249 N.C. App. at 674 (quoting Snead, “a detailed chain of custody for the video need not be shown unless the video is ‘not readily identifiable or is susceptible to alteration and there is reason to believe that it may have been altered’”).
But just to be sure, the proponent might also want to ask a few additional questions to demonstrate the reliability of the downloading or copying process. For example, has the witness ever previously saved or emailed a video clip from this camera? Did it work correctly when she did that? Did she download/save/copy this particular video in the same way? Did she alter it in any way afterward? And then finally, is the video here in court the same one she previously saw and downloaded?
3. Original Required, Best Evidence
Finally, what about the fact that Rule 1002 requires the use of the “original” to prove the contents of a recording? Isn’t the original of a digital video just a bunch of 1’s and 0’s floating in the magnetic aether on a faraway server’s hard drive? Good luck producing that in court.
The proponent has a couple answers to this argument. First, by definition under Rule 1001(3), “if data are stored in a computer or similar device, any printout or other output readable by sight, shown to reflect the data accurately, is an ‘original.”’ So a visual representation of the underlying data (i.e., the video being played) is an original. More importantly, under Rule 1001(4), a copy or download obtained by the witness would be a duplicate created “by mechanical or electronic re-recording,” which makes it admissible under Rule 1003 to the same extent that an original would be, as long as there is no “genuine question” about its authenticity or other reason why it would be unfair to admit it.
Are you seeing these videos in your cases? Are they being successfully contested or routinely admitted? If they are being excluded, is it based on arguments other than the ones discussed here? Please share any helpful thoughts or experiences in the comments.