Article

The Police Journal:

Theory, Practice and Principles
Intelligent evidence: 2017, Vol. 90(1) 55–69
ª The Author(s) 2016
Using open source Reprints and permission:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0032258X16671031
intelligence (OSINT) journals.sagepub.com/home/pjx

in criminal proceedings

Fraser Sampson
Chief Executive & Solicitor, Office of the Police & Crime Commissioner,
West Yorkshire, Wakefield, UK

Abstract
Law Enforcement Agencies (LEAs) are coming to terms with the potency of social media
and Internet-based communication, not solely as an extension of mass communication
but as a phenomenological source of intelligence. One feature of the expansion of
material – particularly that which is openly available to investigators – is the narrowing of
traditional boundaries between information to support lines of activity (intelligence) and
material to be relied on during a criminal trial (evidence). This article addresses the legal
considerations facing LEAs when this concatenation of two different categories of
material occurs and matters of how to reconcile them.

Keywords
OSINT, evidence, social media, intelligence

Introduction
Law Enforcement Agencies (LEAs) around the world are rapidly coming to terms with
the potency of social media and Internet-based communication, not solely as an exten-
sion of their own mass communication (Bruns and Burgess, 2012; Coptich and Fox,
2010; Crump, 2011) but as a phenomenological source of intelligence that for centuries
has been the lifeblood of criminal investigation. One feature of the truly exponential
expansion of accessible material – particularly that material which is openly available to
investigators (Akhgar et al., 2015; Staniforth and Akhgar, 2015) – appears to be the
narrowing of the traditional boundaries between information to support lines of inquiry

Corresponding author:
Fraser Sampson, Office of the Police & Crime Commissioner, West Yorkshire, Ploughland House, George St,
Wakefield WF1 1DL, UK.
Email: [email protected]
56 The Police Journal: Theory, Practice and Principles 90(1)

and activity (intelligence) and material to be deployed and relied on during a criminal
trial (evidence). The question that this article addresses is: what are the legal considera-
tions facing LEAs when this concatenation of two very different categories of material
occurs and how might they be reconciled?

What is OSINT?
The concept of open source intelligence (OSINT) is relatively new for LEAs and is
loosely defined as intelligence collected from publicly available sources that does not
require covert or clandestine methods of collection (Brunet and Claudon, 2015; Price,
2011). The potential of so-called ‘Big Data’ and the analytical tools being developed
alongside it is, for LEAs, phenomenal (Armstrong et al., 2013; Blackman, 2008).
Principal among the many accessible ‘open sources’ used by LEAs are the Internet
and the more popular elements of ‘social media’. There is not space here to rehearse
even the fundamental elements of this area of data processing but for the purposes of
illustration in this article it suffices to adopt Kaplan and Haenlein’s (2010) main
varieties of social media:

1. collaborative projects (e.g. Wikipedia);

2. blogs and microblogs (e.g. Twitter);
3. content communities (e.g. YouTube);
4. social networking sites (e.g. Facebook);
5. virtual game worlds (e.g. World of Warcraft); and
6. virtual social worlds (e.g. Second Life).

Of these, the sources at 2, 3 and 4 have particular relevance in terms of both their
intelligence value and their evidential potential for criminal trials. In addition, the Big
Data capability of LEAs to access data showing, for example, the location of a device
when a photograph was taken or a text sent (Lin, 1998; Seretan et al., 2003; Smadja,
1993) or the temporary and evanescent existence of a digital ‘community coalescing
around a one-off event such as a political rally or pop concert (Beguerisse-Dı́az et al.,
2014) opens up a source of potential evidence that was almost unimaginable at the time
most of our laws of evidence were formulated.
This article focuses on the law and procedure in England and Wales but, while each
jurisdiction will be governed by its own domestic laws, there are nevertheless some
common elements around evidence; there are also some significant overarching provi-
sions within the European Convention on Human Rights (ECHR)1 that will apply to
relevant proceedings in each of the 47 signatory States.2 In the discussion that follows,
the generic principles of evidence and the jurisdiction-specific aspects are considered
along with some examples to illustrate the evidential and procedural issues and the likely
approach of the courts.

OSINT as evidence
Consider the following scenario. A teenage girl makes a complaint to the police that
she has been raped by a friend. Investigators trawling OSINT sources find entries on
Sampson 57

the suspect’s Facebook wall posted by the complainant (Kaplan and Haenlein’s (2010)
fourth variety of social media). Others in their respective Facebook groups can see
these messages, some of which appear to show that the complainant had previously
expressed a liking for the friend and were posted around Valentine’s Day. There are no
replies from him.
The suspect – a 15-year-old boy – is arrested and subsequently charged with rape.
Investigators also find a photo of the complainant in her underwear saved on the
suspect’s Facebook account. Which if any of the OSINT material will be admissible
as evidence at trial? The answer, of course, depends on a number of variables including
which party wishes to rely on it and for what purpose(s). Before highlighting those
variables and assessing how they might be applied by a court, it is necessary to take a
brief look at some basic evidential principles.
Intelligence and evidence are, if not mutually exclusive, then at least substantively
and purposively different. Open Source Intelligence may be gathered for a variety of
purposes, including tracing a suspect, locating a vulnerable missing person or preventing
a planned crime. While LEAs will gather intelligence from a wide variety of sources,
open and closed, there is no requirement – or often even a likelihood – that the product
will be adduced in criminal proceedings.3 If and when it is, the use of such material as
evidence will be open to challenge.4
So, can intelligence ever be relied on as evidence? The answer is simple: yes,
provided the intelligence meets the relevant requirements for evidence of that type.
Aside from the very specific statutory exclusions, such as the intercept intelligence
referred to above, there is no specific reason why intelligence material ought not to
be adduced as evidence. There are however further considerations that need to be
addressed by investigators before relying on such material.

Principles of evidence
In mature legal systems with a developed observance of the rule of law, the rules of
evidence themselves form a substantial body of jurisprudence. Wherever the juris-
diction, legal proceedings properly conducted will essentially involve the presenta-
tion of evidence by the respective parties, either tending to support their case and/or
undermine that of the other(s) during the course of a fair (and usually public)
hearing. And it is the fairness of the proceedings during which that evidence is
presented (or prevented) that is often at the heart of decisions – and appeals – about
the evidence. A defendant in criminal proceedings in democratically developed
jurisdictions will generally enjoy a right to a fair hearing and certain basic entitle-
ments. These would include the right to confront or challenge any witnesses giving
evidence against him or her5 together with some form of qualified protection against
self-incrimination.6
In criminal prosecutions the proceedings include the State who will prefer the
charge(s) against an accused. The subsequent testing of the evidence relied upon by
the parties against the charge(s) will call for decisions to be taken on the facts in issue.
We will revisit this concept with some worked examples below, after considering some
other key features of evidence.
58 The Police Journal: Theory, Practice and Principles 90(1)

Facts in issue, admissibility and weight

In very broad terms two principal questions arise with any material that is going to be
used as evidence: admissibility and weight. Plainly not all potential evidence is admis-
sible,7 not all admissible evidence will be admitted in the hearing and not all admitted
evidence carries the same weight. This is not as complicated as it might appear and the
following illustrations help to clarify the concepts.
In order to be admissible, evidence must be relevant to a fact in issue. The burden of
demonstrating relevance generally falls on the party seeking to rely on it. This can often
be agreed in advance by the parties but will sometimes require a specific ruling by the
judge or tribunal. ‘Relevant evidence’ is essentially material which makes the matter
requiring proof more or less probable (see, for example, DPP v Kilbourne [1973] AC
729). If a fact is not in issue (for example, the consent of a complainant in a sexual
offence) evidence of it is irrelevant and is highly unlikely to be admissible at all (see, for
example, R v Blastland [1986] AC 41).
Therefore if an investigator intends to rely on any evidence – whether intelligence or
otherwise – they will need to identify its ‘materiality’, that is the material must have
some aetiological connection to a fact in issue. This could be the defendant’s motive, the
alleged actions of a witness or the provenance of a document, etc. The need to demon-
strate materiality is particularly prominent in criminal matters where it is not unusual to
have ‘trials within trials’,8 which involve the presentation of argument in the absence of a
jury to obtain a specific legal ruling by the trial judge on whether a particular piece of
evidence can be admitted. The facts in issue in a criminal trial will be established by the
ingredients of the offence(s) with which the defendant is charged and their defence to
those charges. Some evidence will be directly probative of the facts in issue (such as the
defendant’s state of mind in an offence requiring dishonesty) while other evidence may
be circumstantial, that is, evidence of relevant facts from which the existence of matters
in issue may properly be inferred. For example, in a murder trial, the fact that the
defendant delivered a kick to the victim’s head may be a fact in issue; similarly any
preceding demonstrations of enmity from the defendant towards the deceased may also
be key to questions of motive and likelihood that the assailant was the defendant. Matters
of identification or alibi may also be facts in issue and all of these can be particularly
relevant where social media data are available.
Once evidence has been determined to be admissible, the finder of fact (a jury if there
is one, judge/tribunal if not) will have to decide how persuasive it is. Just because it is
relevant does not mean it will be determinative of the whole case. If, for example, the
evidence comes from an unreliable source or if the witness providing it is less than
convincing, these will clearly affect the weight to be ascribed to it. Before moving on to
consider our example, it is necessary to consider a further key principle: that of purpose.

Purpose
If a party wishes to adduce evidence it will need to be for a clear relevant purpose. It is
not unusual for the scope of evidence to be quite narrow. For example, in both England
and Wales and the United States, a statement made in the presence and hearing of the
Sampson 59

accused before they were arrested has historically been admissible, not as evidence of the
truth of the statement itself but as evidence of their reaction to it.9 This ‘purposive’
approach to items of evidence is critical in understanding the uses to which OSINT
material may be put and the limitations or conditions that a court/tribunal may impose
before it is admitted. Evidence might be admissible for more than one purpose. In the
hypothetical murder case referred to above, a recording made by a witness on their
mobile phone and posted on YouTube may, for example, be put forward for the purposes
of proving both the identity of the defendant and the fact that they kicked the victim –
and perhaps that they did so more than once and that there was no other person imme-
diately present. Alternatively, social media data may be adduced by a defendant to
support evidence of alibi.
For some criminal offences (such as the making of threats or insulting comments,10
or fraud) the ‘open source’ material such as Twitter (Kaplan and Haenlein’s (2010)
second variety of social media) may itself be prima facie evidence of the offence; the
purpose in admitting the material would be to prove the ingredients of that offence. In a
trial for assault one fact in issue might be whether the defendant had had any prior
communication with a victim and either the prosecution or the defence may want to
rely on social media (such as Facebook entries or Twitter exchanges) for the purpose of
proving/disproving that fact in issue. Other foreseeable circumstances might include
cases where the prosecution want to rely on social media entries made by a defendant
to show the defendant’s mannerisms, style of writing or other idiosyncrasy (Ormerod,
2016). Alternatively the defence may wish to rely on OSINT to prove, for example in a
homicide, that someone else had been present at the time11 and that they had the same
motive as the defendant. In sum, if investigators intend to rely on any OSINT material
as evidence it is important for them to be very clear about its relevance, intended
purpose, how it relates to any facts in issue in the case and what weight can fairly
be given to it. And the concept of fairness is central to an understanding of the
procedural issues that will arise. Here is why.

Fairness, disclosure and a word about hearsay

In those countries that are signatories to the ECHR, the admissibility of evidence is
primarily a matter for regulation under national law;12 however Article 6(1) requires that
the prosecution authorities disclose to the defence all material evidence in their posses-
sion for or against the accused.13 This duty of disclosure (often strengthened by domestic
legislation such as the Criminal Procedure and Investigations Act 1996 in England and
Wales) is an important element in the evidential use of OSINT and while the rules of
evidence can differ significantly between common law jurisdictions (such as the UK)
and ‘civil law’ jurisdictions (such as those countries whose legal systems evolved from
the Napoleonic Code)14 the overriding requirements of a fair hearing will usually
impose an irreducible minimum level of disclosure. Both fairness and disclosure will
need to be considered if OSINT is to be relied upon in criminal proceedings.
Many, if not all, jurisdictions will also have specific rules about so-called hearsay
evidence and its admissibility. The definition used in England and Wales is that hearsay
evidence is ‘a statement not made in oral evidence in the proceedings that is evidence of
60 The Police Journal: Theory, Practice and Principles 90(1)

any matter stated’.15 While this definition is jurisdiction-specific, the principle is fol-
lowed in other jurisdictions, such as the USA, Canada and Australia.
The rules against admitting hearsay evidence in criminal proceedings have been
substantially relaxed in England and Wales16 and there are a number of statutory ‘gate-
ways’ through which hearsay evidence may be introduced. In a case before the Court of
Appeal17 the elements of a hearsay statement were summarised as:
any representation of fact or opinion made by a person by whatever means;
[including] a representation made in a sketch, photofit or other pictorial form, [;if
]the purpose, or one of the purposes, of the person making the statement appears to
the court to have been—

(a) to cause another person to believe the matter, or

(b) to cause another person to act . . . on the basis that the matter is as stated.18

It can be seen immediately that documents obtained from open sources tendered in
evidence as proof of any matter stated within them will generally meet this definition and
the proposed use of OSINT material arguments as evidence in criminal trials can be
expected to generate argument about hearsay admissibility. In an OSINT context the
rules of hearsay in England and Wales can be seen in two cases where text messages on a
phone were relied upon as both admissible hearsay (R v Leonard [2009] EWCA Crim
1251) and non-hearsay material (R v Twist [2011] EWCA Crim 1143). There is neither
space nor scope here for a trans-jurisdictional comparative analysis of the rules of
hearsay. Suffice it to say that each jurisdiction will have its own rules for the admissi-
bility of hearsay evidence, which, it should be remembered, is not solely a matter for the
prosecution.19 Any admissibility considerations surrounding OSINT material should be
approached by reference to the statutory language and provisions and with the express
view of the prosecutor.

Applying the evidential principles

Returning to our working example of the teenage girl making a complaint of rape by a
friend. Essentially the hypothetical facts in the question are the same as those arising in a
prosecution involving an allegation of rape of a 13-year-old girl by a boy known to her.20
There was – perhaps inevitably given the ages of the parties – an amount of open source
information available to the prosecution and the defence. In particular, the defendant
wanted to put the Facebook messages before the jury; he also wanted to show a photo-
graph of the girl that the defendant said she emailed to him. In his defence statement21
the defendant denied that he had ever had sexual intercourse with the girl. The defence
also sought22 to adduce evidence of a previous complaint by the girl in relation to a
different incident but there was no reference to matters subsequently raised at trial. After
the girl’s evidence via video interview, the defence sought leave to introduce in cross-
examination the photograph that the defendant alleged she had sent to him around about
Valentine’s Day in 2010. The photograph had been taken by the complainant and showed
a ‘selfie’ image in a mirror of her dressed in a bikini or underwear and described as being
‘quite graphic’. The principal questions here in relying on the material as described
Sampson 61

above will be that of relevance, purpose and weight. In T the defence sought to introduce
the photograph in order to prove motive for the complainant’s making of a false com-
plaint of rape against the defendant. However no emails enclosing the photograph were
produced. The defence also sought to introduce a number of Facebook messages that
they alleged had passed between the girl and the defendant. The first message had been
sent some 18 months before the alleged incident and recorded that the girl had added the
defendant as a ‘friend’ on Facebook. There followed a number of ‘chat’ messages on
Facebook but none was shown to have had any response from the defendant and all
showed communication in one direction.
On hearing the defendant’s appeal against conviction the Court of Appeal said that, in
the absence of any evidence of an email accompanying the photograph, the prosecution
had been rightly sceptical as to its provenance as it could so easily have been obtained by
means other than a direct email posting to the defendant (emphasis added). The question
of provenance of OSINT materials – where, when and how they were created, by whom,
for what purpose, who knew about them, how easily they might have been altered and so
on – will be crucial to a determination of their admissibility and weight. There had been
no explanation as to why the photograph had been adduced so late or as to why there had
been no reference in the defence statement to a hostile motive or to the factual basis upon
which such a motive was to be alleged.
Moving on to the specific relevance of the social media material, the question was
whether the photograph had gone to a fact in issue. As the defendant denied that he had
had sexual intercourse at all with the complainant, consent at the time was not in issue23
(and the material had no relevance to it in any event). However, the material could
potentially have gone to the issue of whether the complainant had been ‘interested’ in
the defendant. This very much was a fact in issue and the defendant had claimed not to
have been ‘interested’ in the complainant at all. His defence was that her motive for
making a false allegation had been her affront at his lack of interest. The court held that,
once that relevance had been established, the judge should have allowed the matter to be
put to the complainant and the defendant to give evidence about it.24 The court also held
that the material being raised so late in the day went to the weight to be attached to the
photograph rather than its admissibility, illustrating neatly how the various elements of
evidence summarised above can come together in criminal proceedings involving
OSINT material.
The court held that the fact that the complainant denied having sent the photograph to
the defendant had not resolved the issue and that the court should have heard from the
defendant too. He had wanted to say that he had been sent that photograph by the
complainant and that was a conflict of testimony that the jury would have had to resolve.
If, as was believed by the court, the photograph and questions about it related to a
relevant issue [emphasis added], then it had not been open to the judge to refuse to
allow it merely because the complainant said that she had not sent it. Once it had been
established that the photograph and questions about it related to a fact in issue – namely a
motive for lying – then the judge should have allowed the defence to cross-examine
about the photograph and adduce evidence about it.
The court went on to point out how easily the photograph might have been obtained
from another source, particularly bearing in mind the defendant’s explanation for the late
62 The Police Journal: Theory, Practice and Principles 90(1)

disclosure of the photograph and the Facebook entries: some defect in the hard drive of
the computer used by him. On that matter, the court held that further evidence and
consideration had been required. Again this illustrates a crucial practical area for those
gathering OSINT material if it is to be relied on as evidence in legal proceedings.
To summarise, this case illustrates the key evidential considerations of the OSINT
material, namely: where it came from, who made it, the purposes for which the party
wished to rely on it, the fairness of allowing them to do so, its relevance/materiality to
any fact in issue (including those raised in the defence) and the context in which it should
be considered, the reliability of the witnesses, the weight to be attached to any evidence
once admitted and the technical functioning of the computer on which some of the
material had been processed. All in all this represents a pretty comprehensive illustration
of the issues discussed above.

Some practical guidance

Although they relate specifically to hearsay evidence (as discussed above) the Crown
Prosecution Service guidelines are instructive in underscoring some of the wider
evidential considerations that will concern a court when faced with OSINT materials.
These have been adapted for the purposes of the discussion here and investigators
should ask themselves:

 How much probative value does the material have in relation to a matter in issue,
or how valuable it is to an understanding of other evidence in the case?
 What other evidence has been, or can be, given on the matter or evidence men-
tioned above?
 How important is the matter or evidence mentioned in the context of the case as a
whole?
 How difficult will it be to challenge the material?
 To what extent is that difficulty likely to prejudice the party facing it?
 How reliable does the maker appear to be?
 How reliable does the evidence of the making of the material appear to be?
 In what circumstances was the material made or obtained?

The last three points will affect the weight attributed to any evidence and some-
times its admissibility generally. Whether it be a percipient witness, an admission by a
defendant, a document, a photograph, scientific data or a social media feed, the
provenance of material relied upon and the integrity of the process by which it has
reached the court will be highly significant – and often determinative – of the mate-
rial’s admissibility and weight. The greater the likelihood that the material might have
been easily altered or interfered with, the less likely it is to have any substantial
weight attributed to it. Similarly, where information is orphaned, anonymous or has
no individual willing to testify to its provenance, the less helpful it will be to a court or
tribunal in testing the facts of a case and arguably the less fair it would be to admit it
against one party.
Sampson 63

Having considered the evidential principles, applied them to a specific set of facts and
adapted some published guidance for prosecutors, we can move to consider another set
of circumstances.

Applying the principles again

Following a fatal shooting in a suspected gang-related attack, a police officer conducts
an Internet search and finds a number of photographs of a suspect on a BEBO page. The
suspect (who had been arrested very near the scene some 10 minutes after the attack)
appears to have taken the pictures of himself after he had left prison fairly recently and
they have been digitally placed on the page in a way that suggest he is bragging as a
member of the relevant gang, members of which engage in serious criminal violence. He
is referred to in the material by his ‘street name’ Hustla and there are a few pieces of text
about his coming out of prison. The BEBO page includes a hypertext link to a YouTube
page portraying the gang as violent, although there is no picture of the suspect on the
YouTube page.
What evidential use can be made of this OSINT material in the subsequent criminal
proceedings against the suspect for assisting in the murder?
These were some of the evidential issues that arose in another English case25 and
again they are helpful in understanding how courts will approach some of the evidential
challenges of relying on OSINT material in criminal trials.
In this case the defendant said he had taken the photographs of himself after he had
left prison. Someone had placed the photographs on BEBO portraying him as belonging
to the ‘OC gang’, the members of which engaged in serious criminal violence. The
material referred to the defendant as ‘Hustla’ and associated him with the text ‘Soon
touch road’ and the digit ‘3’. Although the prosecution had no evidence of the IP address
from which the material had been uploaded, the entirety of the BEBO material (com-
prising some 46 separate ‘pages’) was copied for the jury, including a hypertext link to a
YouTube page portraying the ‘OC gang’ as violent. There had been no picture of the
defendant on the YouTube page but it had been downloaded and saved onto a DVD and
shown to the jury.
The trial judge ruled on the admissibility of the material and directed the jury about
(inter alia) its weight. The judge considered the relevant law26 allowing the exclusion of
evidence. Following his conviction the defendant appealed on the basis of both the ruling
that evidence from the BEBO ‘page’ and the YouTube ‘page’ had been admissible and
concerns about the judge’s direction to the jury about that material. The Court of Appeal
held that the material had been inadmissible and that the judge’s direction had not cured
the problems that arose from admitting it.
In the ruling on the BEBO material, the judge had said that, if the defendant had been
the author of all or some of the material on that website then it was plainly admissible. If,
however, the jury concluded that the defendant may not have been involved in the
compilation of that website, they had still been entitled to receive that evidence as part
of the general background to the case [emphasis added].
The defendant had denied any involvement in, or knowledge of, the material, both on
the voir dire and in his evidence before the jury. In his testimony he had said that he did
64 The Police Journal: Theory, Practice and Principles 90(1)

not feature in the YouTube video and had not been involved in the making of it. The
BEBO material had not come from his website and he had not been involved in its
creation. The defendant asserted that he had not known of the website before his arrest,
he had not accessed it and had not had a password to access it. In fact the first time the
defendant had been aware of the material had been when it appeared in the paperwork of
the prosecution. The photographs had shown him at his grandmother’s house and had
been taken shortly after his release from prison several years earlier. He had taken them
with his cousin’s telephone. No one else had been involved in taking them and he had
explained that the word ‘3’ meant ‘free’ from prison, while ‘Soon touch road’ meant that
he would soon be coming home. He had testified that he spelled his street name ‘Hustlar’
with an ‘AR’, not as appeared in the material, and he had taken the jury through other
initials that appeared on the website. He claimed he had not been responsible for what
appeared on the website.
The defendant’s lawyers also argued that both the BEBO and YouTube material were
hearsay evidence and could not be admitted otherwise than through one of the statutory
‘gateways’ (as discussed above).
The court held that it seemed likely that the account holder was representing as fact or
opinion that the defendant was at the time a member of the OC gang. In order for the
material to be admissible the judge had to be satisfied that it was in the interests of justice
to admit it. Similar considerations applied to the YouTube page and there were several
technical submissions about the manner in which the judge had approached the issue of
admissibility.
The court held that, given that the central fact in issue in the case had been whether the
defendant had been innocently at the scene of the shooting, the BEBO material was
potentially very damaging to his case. The court went on to say that, among other things,
the judge had to consider how reliable the maker of the statement had been but had not
identified the maker of the material. Without having an identification of the maker of the
material it was unclear how many different levels of hearsay had been involved.
The trial judge had not considered the reliability of the maker of the statement that the
appellant was a member of the OC gang, and the court cited a previous authority:27
. . . If it appears to the judge that the maker of the statement is unreliable that is a
powerful indication that the statement should not be admitted in the interests of justice.
The court held that, on the facts of the case, the judge should have considered how
reliable the statement had been, that he should also have asked whether the prosecution
could have called the maker of the statement and, if not, why not. There were also obiter
remarks about whether claims (‘bragging’) could amount to evidence of confession
(which would attract its own specific rules for exclusion.)28 While much of this case
turned on the specific application of the relevant statutory provisions of the hearsay laws
in England and Wales and on the judge’s directions to the jury, the approach of the Court
of Appeal helps to illustrate further the evidential considerations of OSINT material used
in criminal proceedings, mainly around relevance to a fact in issue, purpose, reliability
and the issue of fairness.
Before leaving the principles of evidence, there is one final elemental consider-
ation for LEAs when relying on intelligence and that is the manner in which it has
been obtained.
Sampson 65

Obtaining evidence, illegality and breach of process

A core issue regarding the admissibility of material obtained by an LEA in a prosecution
will be the means by which it has been obtained. As has been seen, there is a general
principle of evidence that the interests of justice are paramount: if it would be ‘gravely
prejudicial’ to the defendant to admit the material then, even though technically admis-
sible, that evidence may be excluded.29 In some cases (for instance where evidence has
been obtained as a result of a violation of Article 3 ECHR – the prohibition against
torture, or inhuman or degrading treatment) the violation will render the proceedings as
a whole automatically unfair.30 If the material relied upon has been obtained unlawfully
by the LEA then there are substantial and significant barriers to its being deployed in
proceedings, particularly in criminal trials. The definition and parameters of what
amounts to the ‘unlawful’ obtaining of material are themselves often the source of
considerable dispute and have arisen many times in cases where the material in question
has been obtained in breach of a defendant’s rights under Article 8 of the European
Convention on Human Rights (ECHR).31 If, for example, the LEA obtained OSINT
containing personal data in breach of the appropriate statutory procedure for doing so,
the breach allows the admissibility of the material to be challenged (for example as a
potential breach of Article 8 of the ECHR – Perry v UK (2004) 39 EHRR 76.)
The key element here is the fair administration of justice which, in the view of the
European Court of Human Rights, ‘holds so prominent a place in a democratic socie-
ty . . . it cannot be sacrificed for the sake of expedience’, and the actions of LEAs in
gathering evidence must not, for example, amount to encouragement or incitement of
offences.32 Even where other jurisdictions are involved in the proceedings, if there has
been a deliberate breach of official process by the LEA, there is a substantial risk of the
evidence being excluded.33 Examples of this area of exclusion include cases where an
undercover police informant encourages or induces the defendant to commit an offence
and provides him or her with the means to do so,34 and it makes no difference whether
that encouragement or inducement takes place via the use of social media. While an LEA
can present a defendant with an unexceptional opportunity to break the law of which s/he
freely takes advantage,35 great care will need to be taken by LEAs engaging in online or
social media exchanges with potential suspects if the material is to be relied on in
subsequent proceedings against them. Law Enforcement Agencies must not incite, insti-
gate, persuade or pressurise the defendant into an offence.36 Intelligence gathering that
involves trickery, deception or oppression will be liable to exclusion if the product of it is
going to be relied on as evidence.37 Investigators must also be careful to avoid breaching
statutory regulations protecting communications38 and committing specific offences
relating to communications and data39 during the course of their investigation. However,
some material obtained by non-LEA personnel – such as covert filming40 and material
created by investigative journalists41 or even complainants42 – may be admissible. The
methods used to capture OSINT will therefore need careful consideration and documen-
tation (Liberty, 2011) and practices such as ‘mass data capture’ tactics (such as the use of
web crawler software) may render the material open to challenge. Infringements (such as
unlawful surveillance and interference with a defendant’s communications) may also
amount to a breach of Article 8 ECHR, the entitlement to a private life.43
66 The Police Journal: Theory, Practice and Principles 90(1)

Finally, if the OSINT material has been used for journalistic purposes (as is not
uncommon (see, for example, Poell and Borra, 2011; Russell, 2007) it may attract special
statutory treatment. While material created by citizens acting alone in that capacity would
probably not be protected by the usual statutory provisions enjoyed in England and Wales
by journalists44 and is unlikely to abide by the strictures of journalists’ rules for gathering
and contributing material45, material that has been provided to journalists might. If so, the
issues of compulsory disclosure to LEAs and prosecuting agencies become highly sensi-
tive and are likely to involve questions of the journalist’s substantive rights.46

Conclusion
Intelligence has a different function and purpose from evidence. Both function and
purpose can dictate how the intelligence is gathered, recorded and utilised. While over-
lapping, concepts such as ‘reliability’ are also different in the different context of inves-
tigation and prosecution. The ‘end user’ of intelligence is generally the LEA itself while
the recipient of evidence is a court or tribunal. If intelligence is to be relied on in any
form of legal proceedings it will need to meet the same requirements that the court or
tribunal will demand of evidence in any other form. To improve the prospect of OSINT
being admissible and admitted in criminal proceedings the intelligence gatherer should
ask themselves: what fact(s) will the material be used to prove? How far is it capable of
proving that? Who was the ‘maker’ of the material, in what circumstances did they create
the material and for what purpose? How reliable is the maker of the material and how
reliable would evidence of any necessary supporting statement from them be? How does
the material connect the defendant(s) with a key fact in issue? What contrary open-
source evidence is available to the defendant and how will it be dealt with? How did the
LEA come by the material and what processes did they use to get it? The earlier the
gatherer can address the question of likely evidential use the better the prospects of
identifying and remedying weaknesses and finding alternatives, although it is recognised
that, until an indictment is presented and a defence position put forward to the court, it
will be difficult to know what some of the fact-specific issues of that particular prosecu-
tion will be. However, LEAs should at least consider the issues from the perspective of a
court and at least ask themselves questions about the basic fairness of admitting the
particular material in evidence, questions such as how difficult it may be to challenge the
OSINT and the extent to which that difficulty might prejudice the fairness of the pro-
ceedings against the party facing it.
In a world that relies so unquestioningly on information gathered from open sources it
is all too easy to assume that such information will be accepted in every setting, includ-
ing formal legal proceedings. While some open sources of information are clearly more
dependable than others, the evidential gateways for courts and tribunals are well estab-
lished and jealously guarded – failure to consider them may prove fatal to a prosecution
or related proceedings.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship
and/or publication of this article.
Sampson 67

Funding
The author(s) received no financial support for the research, authorship and/or publication of this
article.

Notes
1. Unaffected by the result of the recent referendum on Great Britain’s continued membership of
the European Union.
2. http://www.coe.int/en/web/conventions/search-on-treaties/-/conventions/chartSignature/3
(accessed 15 April 2016).
3. In the case of covert intelligence obtained via an interception warrant in the UK, not only is its
use in proceedings proscribed by law, but revealing its existence can amount to a criminal
offence: Regulation of Investigatory Powers Act 2000, s.19. See also The Use of Intercept
Evidence in Terrorism Cases, 24 November 2011 House of Commons Library SN/HA/5249.
4. See, for example, Texieira v Portugal (application 44/1997/828/1034).
5. See, for example, Art. 6(1) ECHR.
6. Funke v France, 44/1997/828/1034; see also O’Halloran and Francis v the United Kingdom
[2007] ECHR 545 – the application of Art. 6 in the context of a right against self-incrimination will
depend on the circumstance of each case; Saunders v the United Kingdom [1997] 23 EHRR 313.
7. Despite recent evolution in the rules permitting, for example, hearsay evidence, as seen in the
courts of England and Wales.
8. In England and Wales these are referred to as voire dire.
9. See dicta in R v Hayter [2005] UKHL 6 at 28; also Gaynor (1957–1958).
10. See, for example, www.theguardian.com/uk/2012/may/22/muamba-twitter-abuse-student-
sorry (accessed 16 April 2016).
11. In a case with facts similar to R v Greenwood [2005] 1 Cr App R 99.
12. Schenk v Switzerland [1988] ECHR 17; Heglas v the Czech Republic [2007] ECHR 5564.
13. Rowe and Davis v the United Kingdom [2000] ECHR 91.
14. See Law Society Gazette 11 April 2016: 13–15.
15. For England and Wales per Criminal Justice Act 2003, s. 114(1).
16. Principally by the Criminal Justice Act 2003.
17. Bucknor v R [2010] EWCA Crim 1152.
18. Per the Criminal Justice Act 2003, s. 115.
19. Thomas v United Kingdom [2014] ECHR 1195.
20. T v R [2012] EWCA Crim 2358.
21. A mandatory document in which defendants must set out the basis of their defence to the
indictment brought by the Crown – see Criminal Procedure and Investigations Act 1996, ss 5–
6; Criminal Procedure and Investigations Act 1996 (Defence Disclosure Time Limits) Reg-
ulations 2011; Criminal Procedure Rules, rule 22.4.
22. Pursuant to Youth Justice and Criminal Evidence Act 1999, s. 41.
23. A principal element of the offence of rape (Sexual Offences Act 2003, ss. 1, 74).
24. Following R v F [2005] Cr App R 13.
25. Bucknor, above n. 17.
26. Police and Criminal Evidence Act 1984, s. 78, the general statutory provision allowing the
exclusion of evidence in criminal proceedings.
68 The Police Journal: Theory, Practice and Principles 90(1)

27. R v Musone [2007] EWCA Crim 1237.

28. Police and Criminal Evidence Act 1986, s. 176.
29. See Noor-Mohamed v R [1949] AC 182.
30. El Haski v Belgium (Application no. 649/08) Gäfgen v Germany [2010] ECHR 759.
31. For a detailed analysis of the relevant law relating to EU member states, see ‘Opinion on the
status of illegally obtained evidence in criminal procedures in the Member States of the
European Union’ 30 Nov 2003. Reference: CFR-CDF.opinion3-2003.
32. Khudobin v Russia (application no. 59696/00); Texieira v Portugal (application 44/1997/828/
1034).
33. See R v Quinn [1990] Crim LR 58.
34. Khudobin, above n. 32.
35. Nottingham City Council v Amin [2000] 2 All ER 946.
36. R v Loosely [2001] 4 All ER 897.
37. See e.g. R v Fox [1986] AC 281; R v Alladice (1988) 87 Cr App R 380.
38. See Khan v UK (2001) 31 EHRR 1016.
39. For example, under the Computer Misuse Act 1990, s. 1(1) in England and Wales, when
accessing data in a way that is unauthorised by either the relevant account holder or service
provider (see for example DPP v Bignell [1998] 1 Cr App R 1 and R v Bow Street Metro-
politan Stipendiary Magistrate ex parte United States (No. 2) [2000] 2 AC 216) and also
(2013) Social media and criminal justice. Criminal Bar Quarterly 4, The Criminal Bar
Association.
40. R v Loveridge (200) 2 Cr App R 591.
41. R v Marriner [2002] EWCA Crim 2855.
42. R v Rosenberg [2006] EWCA Crim 6.
43. El Haski, above n. 30.
44. See the Police & Criminal Evidence Act 1984, ss. 11 and 13.
45. See the Journalists’ Code, published by the National Union of Journalists. Available at: www.
nuj.org.uk/about/nuj-code/; see also Clause 47 of the Deregulation Bill.
46. R (on the application of British Sky Broadcasting Ltd.) v The Commissioner of Police of the
Metropolis [2014] UKSC 17.

References
Akhgar B, Saathoff G, Arabnia H, Hill R, Staniforth A and Bayerl PS (eds) (2015) Application of
Big Data for National Security: A Practitioner’s Guide to Emerging Technologies. Waltham,
MA: Elsevier.
Armstrong T, Zuckerberg M, Page L, Rottenberg E, Smith B and Costelo D (2013) An Open Letter
to Washington. 9 December.
Beguerisse-Dı́az M, Garduno-Hernandez G, Vangelov B, Yaliraki S and Barahona M (2014)
Interest communities and flow roles in directed networks: the Twitter network of the UK riots.
Cornell University Library. Available at: http://arxiv.org/abs/1311.6785 (accessed 12
September 2016).
Blackman J (2008) Omniveillance, Google, privacy in public, and the right to your digital identity:
A tort for recording and disseminating an individual’s image over the Internet. Santa Clara Law
Review 49: 313.
Sampson 69

Brunet J and Claudon N (2015) Military and big data revolution. In Akhgar B, Saathoff G, Arabnia
H, Hill R, Staniforth A and Bayerl PS (eds) Application of Big Data for National Security: A
Practitioner’s Guide to Emerging Technologies. Waltham, MA: Elsevier, ch. 7.
Bruns A and Burgess J (2012) #qldfloods and @QPSMedia: Crisis Communication on Twitter in
the 2011 South-East Queensland Floods. Brisbane: ARC Centre of Excellence for Creative
Industries and Innovation, Queensland University of Technology.
Copitch G and Fox C (2010) Using social media as a means of improving public confidence. Safer
Communities 9(2): 42–48.
Crump J (2011) What are the police doing on Twitter? Social media, the police and the public.
Policy and Internet 3(4): article 7.
Gaynor M (1957–1958) Admission in evidence of statements made in the presence of the defen-
dant. Journal of Criminal Law, Criminology and Police Science 48: 193
Kaplan A and Haenlein M (2010) Users of the world, unite! The challenges and opportunities of
social media. Business Horizons 53(1): 59–68.
Liberty (2011) Liberty’s Report on Legal Observing at the TUC March for the Alternative.
Available at: www.liberty-human-rights.org.uk/sites/default/files/libertys-report-on-legal-
observing-at-the-tuc-march-for-the-alternative.pdf (accessed 22 November 2014).
Lin D (1998) Extracting collocations from text corpora. In: First Workshop on Computational
Terminology, Montreal, Canada, pp. 57–63.
Ormerod D (ed) (2016) Blackstone’s Criminal Practice. Oxford: Oxford University Press.
Poell T and Borra E (2011) Twitter, YouTube, and Flickr as platforms of alternative journalism:
The Social Media Account of the 2010 Toronto G20 protests. Journalism 13(6): 695–713
Price E (2011) Library and internet resources for research on terrorism. In Schmid A (ed) The
Routledge Handbook on Terrorism Research. London: Routledge, ch. 7.
Russell A (2007) Digital communication networks and the journalistic field: The 2005 French
riots. Critical Studies in Media Communication 24(4): 285–302.
Seretan V, Nerima L and Wehrli E (2003) Extraction of multi-word collocations using syntactic
bigram composition. In: Proceedings of International Conference on Recent Advances in NLP.
Borovets, Bulgaria, 10–12 September 2003, pp. 424–431.
Smadja F (1993) Retrieving collocations from text: Xtract. Computational Linguistics 19(1):
143–177.
Staniforth A and Akhgar B (2015) Harnessing the power of big data to counter international
terrorism. Akhgar B, Saathoff G, Arabnia H, Hill R, Staniforth A and Bayerl PS (eds)
Application of Big Data for National Security: A Practitioner’s Guide to Emerging
Technologies. Waltham, MA: Elsevier, ch. 3.