Click to Convict? Navigating the Prejudicial Pitfalls of Social Media Evidence in Canadian Courts

By: Austin Lees and Sean White

Introduction:

Since the early 2000s, we have witnessed social media's rapid rise and integration into Canadians' daily lives. It is difficult to imagine a world where we cannot readily access and share information with our friends, families, and professional relationships. Most social media platforms enable this communication by allowing Canadians to share audio, text, and visual content with a few clicks of a button. Police and legal professionals quickly realized the potential of these platforms in their investigations, giving rise to a new form of evidence, referred to as social media evidence.[1] 

One of the significant barriers to social media evidence is that the ownership of an individual's account must be extensively vetted and authenticated.[2] In this blog post, we argue that, in certain circumstances, this authentication process produces substantial prejudicial effects that could outweigh the probative value of the social media evidence. We begin by providing a brief background explaining the two common uses of social media evidence in criminal investigations. After this discussion, we explain the two-stage evidence evaluation framework and describe the distinction between the prejudicial effect and probative value. We then proceed to discuss two substantial prejudicial effects of determining ownership: its invasiveness and the resource-intensive nature of the process. We conclude by offering potential exceptions where the prejudicial effect could be lower depending on the context.

Common Uses of Social Media Evidence:

In the following section, we explain the two common uses of social media evidence in criminal investigations and proceedings. The first category is assessing the credibility of witnesses and victims during proceedings.[3] One of the earliest cases in this area was R v Duran (R v R.D), when the Ontario Court of Appeal (ONCA) wanted to determine a victim's credibility. The accused was charged with sexual assault and interference against the underaged victim.[4] The victim testified that the accused knew that she was underage from posts on her social media page.[5] However, the defense presented evidence that demonstrated that the victim misrepresented her age, meaning the accused reasonably believed she was the age of the majority.[6] The ONCA held that this evidence was admissible and cast doubt on the victim's reliability, impacting the Court's final decision.[7]

The second use of social media evidence in criminal cases is securing admissions from the accused during an investigation.[8] In one scenario, law enforcement will interview the accused about a particular offence and use a public social media entry to induce a confession.[9] In other instances, the Crown may introduce an accused’s public social media entries to prove an element of an offence.[10] For example, in R v Rafferty, the accused was charged with murder of a girl.[11] The Crown presented a list of the accused’s social media posts that demonstrated the accused possessed an essential element of the crime.[12] From this section, we can determine that two dominant uses of social media evidence in criminal investigations and proceedings are assessing the creditability of witnesses and victims, and securing admissions from an accused.

The Authentication Problem and Evaluation Framework

The rise of technology is one of the most notable trends in evidence law and reliable information should not be rejected.[13]  In evidence law, the basic admissibility rule states that evidence must be relevant to a material issue, and determining whether evidence is admissible or whether triers of fact are able to examine certain types of information is one of its most controversial aspects.[14]  Additionally, there are rules implemented to exclude information that is of no use or misleading.[15]  The misleading aspect is particularly challenging in terms of considering the admission of social media evidence because it can be altered, deleted, or fabricated.[16]  In fact, authenticity is one of the most crucial aspects to be considered in Canadian Evidence Law when it comes to social media evidence.[17]  More importantly, the Canada Evidence Act, which classifies social media evidence as “electronic documents,” lays out the admissibility test in section 31.1, stating that: “Any person seeking to admit an electronic document as evidence has the burden of proving its authenticity by evidence capable of supporting a finding that the electronic document is that which it is purported to be.”[18]

This test is particularly crucial, especially as the rise of Artificial Intelligence creates significant trust problems, which can even include the creation of fake court decisions.[19]

Although the threshold for admitting authenticated electronic documents is low, the court will exclude evidence if its prejudicial effect outweighs its probative value. When determining the value of said evidence, a judge considers the strength and believability of the exhibit, as well as cost variables such as fairness to parties and witnesses, practicalities, and the possibility of any distorting effects on the outcome of a case.[20] In R v Dunkley, the Ontario Superior Court (ONSC) decided that the accused's social media posts were inadmissible.[21] 

In this case, the Court referred to Watt’s Manual of Criminal Evidence, specifically regarding how a court can exclude evidence when (1) the prejudicial effect outweighs probative value, (2) the evidence involves an excessive amount of time relative to its value, (3) it is misleading and compromises reliability and, (4) it “involves the needless presentation of cumulative evidence.”[22]  Through the Court’s analysis in Dunkley, most of the evidence was edited or excluded because a number of the Instagram posts were highly prejudicial, irrelevant, or completely unnecessary.[23]  For example, there were Instagram posts of desires to murder unnamed parties, posts portraying sympathy toward the deceased, posts that were considered inflammatory, posts from almost two years after the murder, prejudicial inferences regarding alleged gang involvement, and after-the-fact associations which could have led to misinterpretations about the intentions of the accused.[24] 

Problem One: Invasive Searches

The following two sections discuss two significant prejudicial effects the Crown should consider before authenticating social media evidence. The first effect is the invasive nature of authenticating specific forms of social media evidence. In R v Harris, the accused was charged with the internet luring of an underage youth through Facebook private messages.[25] As part of the case, the Prince Edward Island Superior Court (PESC) wanted proof that the accused was the reliable owner of the Facebook account.[26] The Crown determined ownership by presenting evidence from several private message conversations in the accused’s message history.[27] For example, in an unrelated conversation involving the accused’s friend, the two discussed the accused’s high school, which connected the accused to the account.[28]

The issue with this authentication method is that investigators could violate Section 8 of the Charter of Rights and Freedoms through an unreasonable search and seizure. It may be crucial for the police to review aspects of an accused’s message history with a victim or witness to prove elements of an offence. We agree this aspect is probably not an issue and likely constitutes a reasonable search. However, a problem emerges when investigators analyze conversations that transcend the nature of the initial search, such as messages from other individuals unrelated to the offence. These collateral conversations would likely violate Section 8 of the Charter because they were too invasive and because the messages captured were more detailed than required. 

Problem Two: Resource Intensive in a Strained Legal System

With Canadian courts already experiencing significant quantities of evidence, staffing shortages, and intensive and complicated cases, how will such a strained system have the ability to manage social media evidence on an ongoing basis?[29]  It seems likely that the copious amounts of social media evidence to tender in an already resource strained system may also create unreasonable delays which would also go against the framework established in R v Jordan. If and when this happens, charges might be stayed and produce a negative domino effect in that victims and their families may not see justice while also formulating a negative impression of Canada’s legal system.[30]  Additionally, technology is constantly evolving at a significant pace and one of the challenges defence lawyers dealing with social media evidence in criminal trials now face is ensuring both that they have an adequate understanding of the laws behind social media evidence and that they are “staying abreast of the evolving digital landscape to effectively leverage social media in their criminal defense strategies.”[31]  Again, as noted earlier in the Dunkley case, relevant evidence can still be excluded after weighing in a cost-benefit analysis in that tendering that evidence can involve an excessive amount of time relative to the overall value.[32] 

Social media evidence can also be too resource intensive in other areas apart from criminal law.  For example, in the civil litigation case of Teunissen v Hulstra, the defendant applied to have four years’ worth of the plaintiff’s Facebook postings admitted into evidence, which totalled up to almost 300 pages.[33]  Many of these postings were completely irrelevant and, relying on a principle stating that an entire book’s worth of documents should never be simply admitted as a whole, the Court held that the defendant should rely only on posts that are relevant and, more importantly, that admitting every single Facebook entry over the past four years would “result in lengthy, unmanageable trials, which the rules of evidence are in place to constrain.”[34]  Lastly, it should be noted that any resource intensive aspects should be carefully considered, especially when reflecting on how the COVID-19 pandemic caused the courts to remain idle for months due to health restrictions such as lockdowns.[35]

Conclusion:

In conclusion, social media evidence is the new reality of Canadian jurisprudence, serving important roles in assessing witness credibility and securing admissions from the accused. However, as the use of social media evidence has expanded, the challenges associated with authenticating social media accounts have also grown. This blog post outlined two significant barriers in the authentication of social media evidence: its invasiveness and its resource-intensive nature. We concluded that authentication can be unreasonably invasive when investigators search beyond relevant communications, infringing on individuals' privacy rights under the Charter. Additionally, with an already strained legal system, authenticating social media accounts requires significant time and resources, which can slow down proceedings and risk compromising justice.

Ultimately, we argue that while social media evidence has clear probative value, there are contexts in which its prejudicial effects may outweigh its usefulness. In cases of invasive or disproportionately burdensome authentication processes, exceptions should be considered to mitigate these effects. As social media evidence continues to evolve, Canadian courts will need to balance probative value and prejudicial impact to maintain fairness and uphold the integrity of the judicial process.

[1] Brock Jones, “The Courts Liked Your Post: Assessing Social Media Evidence in Criminal Proceedings " (2015) 62:3 Crim LQ at pages 373-374.

[2] Jones, supra note 1 at pages 378-380.

[3] Jones, supra note 1 at pages 374-376.

[4] R v R.D, 2013 ONCA 343 at para 7.

[5] R v R.D, supra note 4 at para 8.

[6] Ibid.

[7] R v R.D, supra note 4 at paras 45-48.

[8] Jones, supra note 1 at pages 375-377.

[9] Ibid.

[10] Ibid.

[11] R v Rafferty, 2012 ONSC 703 at para 3.

[12] R v Rafferty, supra note 11 at paras 29-34.

[13] David Paciocco, Palma Paciocco, & Lee Stuesser, The Law of Evidence, 8^th ed (Toronto: Irwin Law Inc., 2020) at 11 and 15. 

[14] Ibid, at 13 and 32. 

[15] Ibid, at 13.

[16] Legal Vidhiya: Admin, “Impact of Social Media in the Law of Evidence” (05 August 2023), online: < legalvidhiya.com >.

[17] Wayne Gorman, “Admissibility of Social Media Evidence in Canada: Part 1” (No specific date), Thoughts from Canada – A Court Review Column: Court Review – Volume 57, online: < judges-juges.ca >.

[18] Wayne Gorman, “Effectively Finding and Using Social Media Evidence in Litigation: The Admissibility of Social Media Evidence” (12 May 2022), The Canadian Bar Association, at 10 and 11, online: < judges-juges.ca >.

[19] Bhaskar Chakravorti “AI and Machine Learning: AI’s Trust Problem – Twelve Persisitent Risks of AI that are Driving Skepticism” (03 May 2024), online: < hbr.org >.

[20] R v Martin, 2021 NLCA 1 at para 31; Paciocco, supra note 13 at 47.

[21] R v Dunkley, 2018 ONSC 2741, at paras 1 and 8.

[22] Ibid, at paras 1, 8, 11, and 20. 

[23] Ibid, at paras 24, 61, 69, 70, 71, and 72. 

[24] Ibid. 

[25] R v Harris, 2010 PESC 32 at paras 1-3.

[26] R v Harris, supra note 25 at paras 13-17.

[27] R v Harris, supra note 25 at paras 12-14.

[28] Jones, supra note 1 at pages 369-371.

[29] Kristen Everson, “Long Delays and Collapsed Cases are Eroding Faith in the Justice System, Lawyers Warn” (31 May 2024), CBC, online: < www.cbc.ca >.

[30] Ibid.

[31] Mayes Telles PLLC, “How Social Media Impacts Criminal Cases” (01 April 2024), Mayes Telles Attorneys, online: < www.mayestelles.com >.

[32] Dunkley, supra note 21 at para 11.

[33] Teunissen v Hulstra, 2017 BCSC 2365, at paras 3, 4, and 20.

[34] Ibid, at paras 12, 16, and 18. 

[35] CBC News, “Canada's Backlogged Civil and Family Courts in 'Crisis,' According to Lawyers Group:

Worst Delays are in Ontario, Where it Takes up to 5 Years to Bring a Civil Action to Trial” (10 July 2023), CBC, online: < www.cbc.ca >.