Social Media as Evidence

Social Media as Evidence

New Technology and Trial Practice Workshop

Stanley Hotel, Port Moresby, 18 – 20 March 2019


The Hon Justice Stephen Estcourt AM

Judge, Supreme Court of Tasmania



With 1.54 billion active daily users of Facebook as at the final quarter of last year, lawyers can no longer ignore the phenomenon as a source of evidence in both criminal and civil trials.

I will deal with Facebook in this brief tour of social media as evidence, because it is the largest of all of the many applications, including the Chinese answer to Twitter – Weibo – or “micro-blog”.

In the USA what is called the “Ultimate Social Media Website Interrogatory” asks plaintiffs about 154 individually named possibilities from “About me” through to “Yfrog”, with the question “for each of the websites and/or services listed below identify your username, the email address associated with the account and the approximate date you joined the website or service. If you have not joined a listed website or service, expressly state that you have never joined that particular website or service.”

With Facebook having commenced in 2004, and Myspace the year before, in 2003, I would have thought that the law would by now have been more richly developed in Australia as to its use and misuse. Surprisingly it is not. The best source of reported judicial consideration of Facebook as evidence is Canada.

Facebook as a passive source of evidence

I would recommend that you should start considering Facebook as a potential source of evidence as soon as you receive instructions or are briefed. Possibly your client or the complainant or the victim might have done so even before you were instructed, in which case you would not be concerned with any ethical issues up to that point. Once you receive the instructions though, you will be ethically restrained in what you can do. With that in mind, once the file comes across your desk it will be you who would fail in your duty if you did not properly advise your client as to the availability of publicly accessible material that may cast doubt on a witness’ veracity.

Obviously the very first thing to check would be whether the person has a Facebook account. It is amazing how many people simply fail to turn off the setting “Do you want search engines outside of Facebook to link to your Profile?” When that setting is on, search engines may link to their profile in search results. However, even when this setting is off, while search engines will stop linking to the person’s profile, that may take some time and, in any event, the person’s public profile can always be found on Facebook in every case if people search for the individual’s name.

I am no expert on Facebook security but two things arise out of these observations. The first is that the world is full of people who know less about cyber security than me, and the second is that you should search swiftly before a person can be advised by his or her own lawyers to limit his public profile and maximise security settings.

I think many people would be surprised that their default Facebook settings may allow “everyone” to see their entire page. If they have such settings of course, you could theoretically “follow” them without “friending” them. However the only advantage of doing that is that as a follower their posts would show up in your news feed. The down side is that as a follower you could be detected.

If you find an unprotected account you should immediately take accurate and comprehensive screenshots, making sure to save the images with time and date tags. It goes without saying that you should only be looking, and not making comments of any description.

Taking screenshots can be as simple as using <print screen> and <paste> functions on your computer, or you could use something like Microsoft Paint (a standard application included in all Windows operating systems) so that each web page can then be preserved into a separate image file that can then be converted to a PDF. Whichever method you use, you should have a weather eye to the trial and give thought to the best and clearest way to present the evidence to the court.

Next, a word about what the Americans call “spoliation”. Spoliation, is the destruction of evidence by a person, and would include the destruction of Facebook records (if that is possible). It is potentially a crime if it is done with intent to mislead any tribunal in any judicial proceeding, or to pervert or defeat the course of justice.

Alternatively, the destruction of evidence can be a ground for an application to permanently stay a proceeding. In Palavi v Queensland Newspapers Pty Ltd (2012) 84 NSWLR 523, the plaintiff’s claim was struck out in defamation proceedings in circumstances where she had produced her Facebook records, but admitted destroying her mobile phone, which had contained relevant material. See also Palavi v 2UE Sydney Pty Ltd [2001] NSWCA 264.

It is appropriate for me here, I think, to mention the ethics of advising the deliberate deletion of Facebook material, or the destruction of a mobile phone or computer hard drive. The Queensland Law Society published in Proctor – March2012 an article written by Stafford Sheppard entitled “Dirty Laundry” in which it was noted that legal practitioners should not advise their clients to “clean up” their Facebook or social media pages where there was a likelihood that such material might be required in legal proceedings.

Be aware though that it would not, in my view, having regard to various USA Bar Association rulings I have read, be unethical for a person’s lawyers to advise him or her to restrict their privacy settings to the absolute minimum, or to logout of the account or deactivate it. And it would not be unethical to advise the claimant to take down social media provided that the material was fully preserved by screenshot or printing. (I am sure everyone knows that even if you deactivate a Facebook account it does not go away anyway and that all that is required to reactivate it is to log in using the associated email address and password.)

Facebook as an active means of surveillance

If a person has left his or her privacy setting as to who may view their posts as “everyone”, or even sometimes just “friends of friends”, you may be able to view ongoing text and images without difficulty. The more common scenario though is likely to be that you have been able to see profile material publicly displayed but cannot see existing and ongoing posts in the private section of the account.

It is important to note that all Facebook “cover page” photographs are “public view”, irrespective of your security settings.  This is the Facebook default.  Hence, even if an individual has locked down their account as far as is possible, friends can still “like” and “comment” on their cover photo, and anyone can right click and open the photo and see the likes and comments. Anyone can then view the profiles of those people on which the individual may appear in photographs or comments.

If it is the case that you have been able to see profile material publicly displayed but cannot see existing and ongoing posts in the private section of the account, then you would need to make a case for discovery or subpoena of the private part of the account based on what you have from the public section. I will deal with that shortly.

There are no laws of which I am aware that would prevent you simply “friending” a person of interest or a “friend of a friend” of that person. However, the relevant consideration is that of your ethical obligations as a legal practitioner.

Rule 3 of the Professional Conduct Rules 1989, by way of example, provides that a solicitor must not engage in conduct which is dishonest or unprofessional, or might bring the legal profession into disrepute. Rule 4 provides that a lawyer shall not attempt to further his client’s case by unfair or dishonest means, and Rule 17 provides that a prosecuting lawyer shall not seek to obtain a conviction by improper means. It would in my view be a breach of any or all of those rules as plainly dishonest and unethical to “friend” a claimant or a claimant’s friend under a false name.

Doing the same under a real identity throws up a more difficult consideration. Many ethical rulings by disciplinary bodies, such as my local Tasmanian ruling which was recently promulgated in a practice guideline, make it clear that a lawyer who is acting on behalf of a party in any contentious matter must not communicate or confer directly with an opponent’s client, except in certain specified circumstances that are not relevant for present purposes.

In my view that does not leave room for a friend request, either by you or by your agent at your request.

A Philadelphia lawyer would be expected to know the answer to this ethical dilemma, surely? Well, the Philadelphia Bar Association considers it necessary to disclose the true identity and purpose or intent of the friend requestor.

The New York Bar Association, on the other hand, considers the disclosing of the true identity of the requestor is all that is necessary, and the fact that that person is, in essence, the agent for the opposing litigant need not be disclosed. The ruling is “while there are ethical boundaries to such ‘friending’ in our view they are not crossed when an attorney or investigator uses only truthful information to obtain access to a website, subject to compliance with all other ethical requirements”.

I doubt that the Legal Profession Board in my State of Tasmania would take a similar view to the New York Bar Association, but the question obviously remains unsettled.

In another recent ethical ruling, the Philadelphia Bar Association Professional Guidance Committee addressed the issue of whether an attorney could direct an investigator to “friend” a claimant. The Committee concluded that doing so would be inherently deceitful and unethical, even if the investigator used his or her own name.

The Committee rejected the contention that obtaining access to a Facebook page was no different to conducting covert video surveillance, saying:

In the video situation, the videographer simply follows the subject and films him as he presents to the public. The videographer does not have to enter a private area to make the video.”

The Committee noted that it would be clearly improper for the videographer to pose, for example, as a utility worker to gain access to someone’s home.

Obtaining discovery in civil proceedings

Rule 6 of the National Court Rules 1983 helpfully defines the word “document” as including “any record of information, whether legible by eye or not”.

I doubt that anyone would seriously argue that a Facebook page on a computer containing writing and images and possibly sounds was not a document for the purposes of discovery in civil proceedings under Order 9 of those Rules. I note that the amended definition of the word “document” inserted in s 1 of the Evidence Act 1975 by s 1 of the Evidence (Amendment) Act  2016 certainly covers the situation. The real question is how do you go about obtaining discovery of the images and text in the private profile of a claimant’s Facebook page?

There are at least two possibilities that might rebut a suggestion of mere “fishing”.

The stronger of the two arguments is where you have been able to locate relevant photographs on the claimant’s public profile. In Murphy v Perger [2007] OJ No 5511 (Ont SCJ), Rady J ordered production of the plaintiff’s entire Facebook page one month before trial on the following basis:

It seems reasonable to conclude that there are likely to be relevant photographs on the site for two reasons. First, is a social networking site where I understand a very large number of photographs are deposited by its audience. Second, given that the public site includes photographs, it seems reasonable to conclude the private site would as well.”

The other possibility which arguably was endorsed by Brown J in Leduc v Roman (2009) 308 DLR (4th) Ontario Superior Court, relying on Murphy v Perger, is the contention that general evidence about how Facebook works and the nature of the service it offers is sufficient to allow a court to infer the likely existence of relevant photographs on the claimant’s private profile.

In support of such an argument the proposition could be advanced that Facebook, given its all-pervading presence in the lives of 1.5 billion humans, makes it an especially important class of material to personal injury litigation.

In Nucci v Target Corp 2015 WL 71726 Fla Dist Ct App Jan 7, 2015, the Appeals Court, Gross J said at 7:

“In a personal injury case where the plaintiff is seeking intangible damages, the fact-finder is required to examine the quality of the plaintiff’s life before and after the accident to determine the extent of the loss. From testimony alone, it is often difficult for the fact-finder to grasp what a plaintiff’s life was like prior to an accident. It would take a great novelist, a Tolstoy, a Dickens, or a Hemingway, to use words to summarize the totality of a prior life. If a photograph is worth a thousand words, there is no better portrayal of what an individual’s life was like than those photographs the individual has chosen to share through social media before the occurrence of an accident causing injury. Such photographs are the equivalent of a ‘day in the life’ slide show produced by the plaintiff before the existence of any motive to manipulate reality. The photographs sought here are thus powerfully relevant to the damage issues in the lawsuit.”

In other words the argument is that a person’s life as depicted on Facebook is as important as his or her evidence at trial as to his or her activities of daily living and, given that the person could be cross-examined as to his or her Facebook posts and called upon to produce them during the trial, fairness dictates that pre-trial discovery should be ordered.

I should add that Nucci (above) is a helpful case in relation to the framing of interrogatories and the tailoring of associated requests for discovery.

Proving Facebook posts

The last issue I wish to deal with is often considered the most difficult. In truth it is not.

Facebook evidence is successfully used every day in criminal and civil trials around Australia in order to disprove persons’ accounts of their activities and as admissions or other proof of guilt of crimes such as identification. The cases are rarely reported of course. A couple of examples that have found their way into the reports are Frost v Kourouche (2014) 86 NSWLR 214 and Munday v Court (2013) 65 MVR 251.

How do you prove a document which comprises a screenshot or a photograph or printout of a Facebook post or page? You simply tender it. If it is relevant it is admissible. The question is one of relevance. It is not necessary to prove the authenticity of such a document before it can be used in evidence.

The judgment of Federal Court judge Perram J in Australian Competition and Consumer Commission v Air New Zealand Ltd (No 1) [2012] FCA 1355 was not concerned with Facebook posts but with business records.  However his Honour’s decision at [92]–[93] is, with respect, a masterpiece of judicial logic and applies equally to Facebook evidence. It is a lengthy passage but it warrants setting out in full. His Honour said:

“92       It is useful to begin with some basic propositions:

1    There is no provision of the Evidence Act which requires that only authentic documents be admitted into evidence. The requirement for admissibility under the Act is that evidence be relevant, not that it be authentic.  On some occasions, the fact that a document is not authentic will be what makes it relevant, ie, in a forgery prosecution.  In other cases, there may be a debate as to whether a particular document is or is not authentic, for example, a contested grant of probate where it said that the testator’s signature is not genuine.

2    In cases of that kind, the issue of authenticity will be for the tribunal of fact to determine.  In cases heard by a judge alone, this will be by the judge at the time that judgment is delivered and the facts found. In cases with a jury, it will be the jury.

3    The question of what evidence will be admitted is a question of law for the tribunal of law, which will be the Court.

4    Since authenticity is not a ground of admissibility under the Evidence Act, the issue of authenticity does not directly arise for the tribunal of law’s consideration at the level of objections to evidence.

5    What does arise for its consideration is the question of relevance under s 55.  If the evidence is relevant it is admissible: s 56. It will be relevant under s 55 if the evidence is such that ‘if it were accepted, [it] could rationally affect (directly or indirectly) the assessment of the probability of the existence of a fact in issue’.

6    The question of a document’s authenticity is relevant only to the tribunal of law’s consideration of relevance under s 55.  It has no other role.

7    In that inquiry, the question for the tribunal of law is not whether the document is authentic but whether receipt of the document could, to paraphrase s 55, rationally affect the assessment of the probability of a fact.

8    If there is raised a question about the authenticity of a document (and assuming that, if authentic, it would otherwise be relevant to an issue) then there will be an issue in the proceedings about its authenticity.  This will be a question for the tribunal of fact to resolve, if the document is admitted.

9    The question for the tribunal of law, by contrast, will be whether the document is relevant to a fact in issue under s 55.  That is, the question will be whether the document can rationally affect the assessment of the probabilities of the fact, including its authenticity.

10  What materials may be examined in answering this question?  The answer is provided by s 58:

58        Inferences as to relevance

(1)        If a question arises as to the relevance of a document or thing, the court may examine it and may draw any reasonable inference from it, including an inference as to its authenticity or identity.

(2)        Subsection (1) does not limit the matters from which inferences may properly be drawn.

11  The position then is clear.  In answering the only question before the tribunal of law – relevance – the tribunal may examine the document to see what may be reasonably inferred from it (s 58(1)).  It may also examine other material (s 58(2)).

12  The tribunal of law does not find that the document is authentic.  It finds that there is, or there is not, a reasonable inference to that effect and hence that the document is, or is not, relevant.  If there is a reasonable inference that the receipt of the document will rationally affect the probability of a finding of fact, then the matter may to go to the tribunal of fact which will then determine at the end of the trial whether the document is authentic and whether the fact is proved.

13  At no time does the tribunal of law determine that the document is or is not authentic because this is not a question for it.  It may, however, determine that no reasonable inference to that effect is open and thereby conclude that it is not relevant

14  In deciding relevance (ie whether the tribunal of fact could reasonably infer that the document (otherwise relevant) was authentic), the tribunal of law is explicitly authorised by s 58(1) to ask what inferences as to authenticity are available from the document itself.  That is what s 58(1) says.

93        It will follow that AirNZ’s submission that ‘no inference as to authenticity can be drawn from the face of these documents’ ought to be rejected. In determining a relevance objection, that is precisely what s 58(1) permits.” (My italics.)

In a nutshell, when you tender a Facebook screenshot or printout no question of its authenticity arises as a threshold question. The only question at that stage is relevance. At no time does the judge as the judge of law determine that the document is or is not authentic because that is not a question for him or her. The question for the judge as the judge of law is only relevance. The question of authenticity is for him or her, after the document has been admitted into evidence, and that is for him or her as the judge of fact.

He or she may, however, as the judge of the law, determine that, on examining the document, no reasonable inference as to authenticity is open, and thereby conclude that it is not relevant.  As an extreme example, if the asserted Facebook post looks more like a family photo album or a school yearbook, then no reasonable inference as to authenticity could possibly be open, and the document would therefore be ruled not relevant and thus not admissible.

These considerations are not dependent upon the particular provisions of the Australian Commonwealth Evidence Act 1995. They would be valid considerations at common law. But, having said that, there is nothing in my respectful opinion in the Evidence Act 1975 (PNG) or the Evidence (Amendment) Act 2016 (PNG) that is inconsistent with Perram J’s approach and much that supports it.

Section 67C of the Evidence (Amendment) Act does refer to the burden of proving the authenticity as lying with the party proposing to give the record in evidence, but that section is referring to the ultimate factual burden of proof and is not seeking to make authenticity a pre-condition to admissibility. If it were otherwise, a forged electronic document could not be proved because it was not authentic.

In deciding the question of relevance, that is, in deciding the question of law as to whether the tribunal of fact could reasonably infer that the document was authentic, the tribunal of law is explicitly authorised by s 67A(3) of the Evidence (Amendment) Act to draw any reasonable inference from the form and content of the electronic record itself. If it looks like a duck and it walks like a duck and it quacks like a duck, then it is sufficiently authentic to be relevant, and thus admissible. In that respect s 67C is not unlike s 58 of the Evidence Act 1995 (Cth). So, if it looks like a Facebook post in form and content, then it has sufficient authenticity to make it relevant and admissible. The ultimate question of authenticity will be decided by the judge as a matter of fact, and s 67C of the Evidence (Amendment) Act dealing with weight has work to do there.

So, you could in theory tender a print of a screenshot in opening your case by handing it up. If it looks like a Facebook page in the claimant’s name and containing the claimant’s photograph as a profile picture or cover photo, then that should be enough to have it admitted. I pause to note that a print of a screenshot or a print of a photograph of the screen image should meet the description of a “printout” in s 67B of the Evidence (Amendment) Act. But even were it otherwise that section would not prevent proof by such means.

On the other hand, you would, I think, often wait to tender the printout through a witness in evidence-in-chief, or in the ordinary course of cross-examination if the person said to have created the post gives evidence.

As to the ultimate question of fact, if a person in his or her evidence denies that the post is his or hers, or claims that it is not genuine, then the issue will play out like any other disputed issue of fact. If the person denies he or she posted it, then that claim will be tested by cross-examination. “Who had access to your account? How was your account hacked? Who knew your password? When was it hacked? What about the posts either side of that post?” If the tribunal is a judge alone, then he or she will decide the question of authenticity and the weight to be given to it (see s 67J of the Evidence (Amendment) Act), when giving his or her final decision.