From Situation Awareness to Action: An Information Security
Management Toolkit for Socio-technical Security Retrospective and
Prospective Analysis
Jean-Louis Huynen and Gabriele Lenzini
SnT - Interdisciplinary Centre for Security, Reliability and Trust, University of Luxembourg, Luxembourg, Luxembourg
{firstname.lastname}@uni.lu
Keywords:
Socio-technical Security, Information Security Management and Reasoning, Root Cause Analysis.
Abstract:
Inspired by the root cause analysis procedures common in safety, we propose a methodology for a prospec-
tive and a retrospective analysis of security and a tool that implements it. When applied prospectively, the
methodology guides analysts to assess socio-technical vulnerabilities in a system, helping them to evaluate
their choices in designing security policies and controls. But the methodology works also retrospectively.
It assists analysts in retrieving the causes of an observed socio-technical attack, guiding them to understand
where the information security management of the system has failed. The methodology is tuned to find causes
that root in the human-related factors that an attacher can exploit to execute its intrusion.
1 INTRODUCTION
Even a ‘secure’ system can turn out to be vulnerable
when attackers target not the system’s technical secu-
rity mechanisms e.g., the cryptographic protocols, but
the system’s users.
In such situations, security cannot be achieved
purely by technical solutions because of its depen-
dency on the non-technical qualities of the system,
such as the system’s usability and the system’s func-
tional design. Thus, achieving security must consider
human-related factors, like the cognitive and psycho-
logical traits that drive human behavior and the peo-
ple’s abilities to interact with information and com-
munication technology.
Failing to understand this holistic complexity
leads to the deployment of systems which are left at
the mercy of attacks of socio-technical nature. And
there is no rhetoric in this warning: the impact of
such attacks is already huge. According to the Ver-
izon’s 2015 Data Breach Investigation Report that re-
ports on the 80,000 security incidents that occurred in
2015, people is ‘the common denominator across the
top four patterns that accounts for nearly 90% of all
incidents’ (Brumfield, 2015). These figures call for a
better understanding of the impact that humans have
This research received the support of the Fond National
de la Recherche (FNR), project Socio-Technical Analysis of
Security and Trust (STAST) I2R-APS-PFN-11STAS, and of
the SnT/PEP-security partnership project.
on security, and prove the need of measures more ef-
fective in reducing the risk of socio-technical attacks.
It is better to clarify that we are not claiming that
organizations are unaware of or that underestimate se-
curity risks. Companies invest in Information Secu-
rity Management (ISM) to keep attacks under control.
They assess risks regularly, and implement measures
that they deem appropriate. But in contexts where the
human behavior is significant for security, predicting
the effect of mitigations only reasoning on technical
grounds may result in a false sense of security. For in-
stance, a policy that force users to change passwords
regularly, supposedly protecting from password theft,
may nudge users to start using easy-to-remember
pass-phrases (Adams and Sasse, 1999) which are also
easy to guess. A policy that seems successful may be
effective only thanks to ‘shadow security’ (Kirlappos
et al., 2014), a spontaneous behavior that people take
in opposition to an otherwise ineffective policy.
So, what reasoning can one resort to unveil a sys-
tem’s socio-technical weak points and what remedia-
tion can one apply to effectively strengthen security?
Addressing this question requires a change of per-
spective. Human users and the untangled factors that
an attacker can manipulate to exert influence on peo-
ple’s behaviours should be included in the security
best practices, analysis, and mitigation strategies.
This seems not to be such a revolutionary thought
at least when we look at other disciplines. In safety,
incident analysts follow practices to achieve similar
Huynen, J-L. and Lenzini, G.
From Situation Awareness to Action: An Information Security Management Toolkit for Socio-technical Security Retrospective and Prospective Analysis.
DOI: 10.5220/0006211302130224
In Proceedings of the 3rd International Conference on Information Systems Security and Privacy (ICISSP 2017), pages 213-224
ISBN: 978-989-758-209-7
Copyright
c
2017 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
213
Figure 1: The PDCA process in the ISO 27001:2005 with the OODA loop for incident response proposed by Schneier.
goals i.e., understanding the root cause of events that
involve human errors. Here, it is common to adopt
a waterfall view of the universe (Anderson, 2008)
where top-down approaches are followed to find out
the causes of unwanted events, where bottom-up ap-
proaches are employed to foresee the negative con-
sequences of a design choice, and where the identi-
fication of the different factors that foster an adverse
outcome is used to set priorities and to optimize the
efforts in defining and applying remedies.
In principle, such practices could be applied in se-
curity as well. An organization following those prac-
tices could improve its situation awareness and orient
better its efforts in mitigating risks. Unfortunately, as
we explain in § 4, applying in security off-the-shelves
methods conceived for safety does not bring the de-
sired result. We need to rethink and adapt them.
Suspending for a moment the discussion about the
character and the difficulties of this adaptation, we be-
lieve that such a migration is possible, and this paper
discusses and implements a way to do it. We refocus
the study of security by centering on the impact that
users have on a system, and we propose our method-
ology as part of the ISM life cycle.
Aim and Contribution. Elaborating on our previ-
ous work (Ferreira et al., 2015), we discuss the need
of a root cause analysis in security (see §2 and 4) and
present a two-way methodology along with a tool to
analyse the impact of human users on security (see
§5). Used retrospectively, the methodology helps an
analyst investigate a security incident by clarifying
how the attack could have pushed users to perform
hazardous actions. Used prospectively, it supports an
analyst investigate a system’s resilience against spe-
cific socio-technical attacks and threat models so re-
vealing vulnerable socio-technical factors that could
help an attacker in its malicious enterprises.
This capability to identify the social and the tech-
nical factors that may have contributed to the suc-
cess of an attack is inspired by Root Cause Analy-
sis (RCA) techniques found in safety, in particular in
one called Cognitive Reliability and Error Analysis
Method (CREAM). In this work, we augment it with
subsidiary capabilities (see §4), the most important
of which is the ability to generalize the knowledge
that an analyst gains inspecting a security intrusion.
This knowledge is structured in such a way that it can
be reused to analyse another system and to identify
its socio-technical weaknesses. Levering on (Ferreira
et al., 2015), we still call the methodology Cogni-
tive Reliability and Error Analysis Method for Socio-
Technical security (S·CREAM).
We also developed a tool to assist analysts follow
our methodology, which we call S·CREAM Assistant.
We exemplify its use in the analysis of the security
of a One Time Password (OTP) solution: the Yubikey
USB security token (see §6).
Our tool is meant to be accessible on-line. Its rea-
soning process and its knowledge base of vulnerabili-
ties are designed to be available to anyone who wishes
to integrate our methodology in an ISM system. Be-
sides, its knowledge base is designed to grow as more
retrospective analyses are performed. However, at the
moment, S·CREAM and its tool are still in a proof-
of-concept phase so in §7 we discuss limitation of the
current version and what we need to do to make it a
fully operational.
2 ON INFORMATION SECURITY
MANAGEMENT
There are different processes that organizations can
adopt in order to structure their ISM efforts. The
ISO 27001:2005 standard (International Organiza-
tion for Standardization, 2005) recommends organi-
zations to follow a cycle called Plan, Do, Check, Act
(PDCA) (Ishikawa and Ishikawa, 1988). The cycle
guides organization in assessing information security
risks and formulating security policies, and the mean-
ing of the cycle’s steps is as follows:
Plan: list assets, define threats, derive the risks
posed by these threats on the assets, and finally
define the appropriate controls and security poli-
cies needed to mitigate the risks;
ICISSP 2017 - 3rd International Conference on Information Systems Security and Privacy
214
Do: conduct business operations while imple-
menting the security controls and enforcing the
security policies defined in ‘Plan’;
Check: check that the ISM system is effectively
mitigating the risks without impeding the business
operations i.e., detect defects in the ISM system.
Act: correct to the defects identified in ‘Check’.
In the PDCA cycle is the ‘Do’ phase where the or-
ganization operates. This is also the phase where at-
tacks can occur and, subsequently, where Incident Re-
sponse fires up. Schneier, in (Schneier, 2014), argues
that Incident Response can also be seen as a loop,
called Observe, Orient, Decide, Act (OODA) (Boyd,
1995), in which organizations engage in four steps:
(i) Observe: observe what happens on the corporate
networks. This is a phase of collection of data in
which, for instance, Security Information and Event
Management (SIEM) systems are used; (ii) Orient:
make sense of the data observed within its context of
operations also by processing the pieces of informa-
tion gathered from previous attacks and threat intelli-
gence feeds; (iii) Decide: decide which action should
be carried out to defend against an attack, at best of
its current knowledge; (iv) Act: perform the action
decided earlier to thwart the attack. Figure 1 shows
the PDCA cycle with the OODA in its ‘Do’ step.
Organizations engaged in such a process usually
implement two types of analyses: a prospective anal-
ysis in the ‘Plan’ step to predict how their security
measures are expected to protect their assets; and,
usually after an attack, a retrospective analysis in the
‘Check’ step to identify the reasons of why their ISM
has failed.
Shortcoming of Current ISM Processes. The ex-
istence of methodologies and tools to assess risks
(e.g., OCTAVE Allegro (Caralli et al., 2007)) help
company implement a PDCA process as we de-
scribed, but there is no method that help them foresee
what impact planned-checked-acted mitigations will
have on the system’s actual security. Indeed, intro-
ducing a security control to contain a risk (e.g., in the
Act’ step) may introduce new socio-technical vulner-
abilities because of unforeseen interactions between
the newly introduced controls and the organization’s
employees, business processes, and existing security
measures.
Additionally, in the case of a security breach, there
is lack of methods (in the ‘Check’ step) encompassing
human-related aspects of security that could be used
to identify the root cause of ISM systems failures.
The operational phase (i.e., the ‘Do’ step) is also
not tuned to consider socio-technical aspects of se-
curity. Indeed, if organizations use Security Infor-
mation and Event Management sytems in their Se-
curity Operations Centers to monitor their networks,
they don’t really consider the human-related aspects
of their operations. For instance, when an company
is hit by a ‘fake president’ scam —a fraud consist-
ing of convincing an employee of a company to make
an emergency bank transfer to a third party in order
to obey an alleged order of a leader, the president,
under a false pretext— the organization can defend
itself by focusing on the technical aspects of the at-
tack (e.g., by blocking connections to a ranges of IPs).
However, the social aspects of the attacks (i.e., that
people does not recognize the phishing and falls for
it) remain, while the attacker can easily adapt to the
additional security controls and persist in using the
same social engineering tricks i.e., phishing people
but in a different kind of scam. Security practition-
ers mainly rely on user awareness campaigns to cover
the social side of these family of attacks but we claim
that identifying the reasons why employees fall for
the scams would be more effective. Indeed, it could
be that the phishing campaign exploits loopholes in
the interactions between the organization’s security
policies and organization’s business processes (hence
employee’s primary goals). Identifying clearly these
reasons could help propose additional remedies that
focus not on the technical, but on the social aspects
of an attack. These new remedies ought to cause the
attacker the hassle of adapting its behaviour and its
attack instead of only fiddling with technical details.
We believe that this strategy could prove beneficial to
the defense as it forces the attacker to climb up the
‘pyramid of pain’ (Bianco, 2014).
Another deficiency that a ISM may suffer is the
lacking of ways to learn from past failures in applying
the remediation strategies. Indeed, without a proper
way to determine the root causes for past failures, or-
ganizations are doomed to repeat ill-considered deci-
sions and lose energy correcting these.
For all these reasons, we believe that security
practitioners should have access to additional meth-
ods that could help pondering the consequences of
their security choices, the technical and the social
alike, to prevent the recurrence of security failures.
3 RELATED WORK
Among the numerous of works on Root Cause Analy-
sis (RCA) in safety incidents we comment only those
in relation to the information security domain. Those
which we have found in the literature use RCA in spe-
cific situations: the process they follow is not flexible
From Situation Awareness to Action: An Information Security Management Toolkit for Socio-technical Security Retrospective and
Prospective Analysis
215
Figure 2: The process usually used for a Root Cause Analysis.
enough to work in other and in general contexts. For
instance, Coroneo et al. (Cotroneo et al., 2016) imple-
ment a strategy for the automated identification of the
root causes of security alerts, but what their offer is a
fast algorithm to timely response to intrusions and at-
tacks which is not applicable to analyse general secu-
rity incidents. Another example of highly specific ap-
plication is a work that customize the search for root
causes in software failures (Kasikci et al., 2015). It
supports software testing with an automated debug-
ging that provides developers with an explanation of
a failure that occurred in production but, still, this is
not a solution that can be generalized to other security
incidents.
A completely different category of works, which,
again, we represent here by picking a single work
in the class i.e., (Schoenfisch et al., 2015), is
that proposing more efficient RCA’s reasoning algo-
rithms. Works in this class to not relate with under-
standing socio-technical causes. Rather they focus on
improving the performances of existing methodolo-
gies but not on extending them to work in security.
Our search for related work seems thus more in-
sightful not in what it has found but rather in what
has not been found, that is, for the lack of works at-
tempting to migrate in security well established RCA
methodologies with the aim of helping as widely as
possible security practitioners. At the time of writing
(October 2016), we have found no significant and re-
lated article that address this problem, nor have we
found RCA in security when human are involved.
Of course there is a plethora of research that stress
the need of keeping the human in the security loop
(see e.g., (Noureddine et al., 2015; Beautement et al.,
2016)) as there are plently of works in field of socio-
technical security, usable security, human factors in
security, and similar topics. For reason of space, this
list is too long to be considered here but, in the best of
our understanding, we were not able to find in those
works methodologies that could help security practi-
tioners in a RCA of socio-technical security incidents.
In summary, our analysis of the state-of-the-art
shows that either we were unable to find representa-
tive pieceworks of research that relate with what this
paper presents or the research questions and the chal-
lenges that this paper tries to address are fact original.
4 RCA IN SAFETY & SECURITY
In safety, the objective of a RCA process is to iden-
tify the cause(s) of an incident. The process will help
gain the knowledge necessary to operate on the fac-
tors that have caused the event and to introduce solu-
tions (e.g., re-designed interfaces or guidelines to op-
erate the system) to impede the event from happening
again.
Following a RCA process typically requires four
steps (see Figure 2): (i) Data collection and inves-
tigation, to produce a description of the incident;
(ii) Retrospective analysis, to inspect the incident in
search for the causes that have triggered the incident
but how the analyst conducts this investigation and
how much the outcomes rely on the analyst’s expe-
rience depends on the specific RCA technique used;
(iii) Recommendations generation, to produce a list
of recommendations whose goal, when implemented,
is to avoid that the incident reoccur; (iv) Recommen-
dations implementation to obtain a safer system, free
from the caveats that caused the incident.
Following this process, an incident’s causes are in-
vestigated thoroughly even when it can be attributed
to ‘human error’. In his accident causation model,
Reason (Reason, 1990) shows that ‘human errors’ are
active failures that, when combined with latent fail-
ures, can transform a simple hazard into an accident.
Thus, it is rare that a person, despite liable for an acci-
dent, is blamed as a conclusion of step (ii); rather the
root cause is found in the complex interplay among
the human, the system, and the context where human
and the system operate, elements that are considered
fertile ground for ‘human errors’ to happen.
4.1 From Safety to Security
Such a way to look at users as potential victim of a
system design’s deficiencies is advisable also in secu-
rity. Too often, human errors in security are seen as
inevitable, the end of the causative chain. In the An-
nual Incident Reports 2015, by ENISA, published in
September 2016, human errors are pointed out be the
‘root cause category involving most users affected,
around 2.6 million user connections on average per
incident’ (ENISA, 2016). The conclusion that hu-
mans are at the root of all such incidents is worrisome
but not helpful. It does not suggest how to improve se-
curity without removing the humans. And in the lack
ICISSP 2017 - 3rd International Conference on Information Systems Security and Privacy
216
Figure 3: S·CREAM’s overall process.
of a comprehensive understanding of the reasons why
systems and processes allow humans to be induced in
security critical errors through which the system is be-
ing attacked, the problem of reducing human-caused
insecurity incidents remains.
That said there are two neat differences between
the RCA currently used in safety and a potential (ret-
rospective) RCA as it should be used in security and
the they are both about defining when to stop the
search for causes. First, the search for root causes
should not end to the attacher who, in security, is the
obvious root of all evil. Second, when humans seem
to be responsible of the incident, the search should
not stop and point the user as the cause of the incident
either, at least, not without looking also for the trig-
gers of human behaviour that the system could have
left to the control of the adversary. Such quest may
reveal that the blame is, at least in part, on the sys-
tem and not on its users. An example is when users
click on poisoned links. Many security policies utter
that clicking on links is a bad habit, but users, who
are daily stormed by trusted and untrusted mails most
of them carrying links, hardly can discern foes from
friends. So perhaps the cause should be looked in the
system’s failing to authenticate an email’s source or
in the reasons why people fail to recognize friends
from intruders that ‘pretext’ to be friends but not in
the ‘users’ as such.
The retrospective analysis of the RCA in safety is
not the only methodology that has a potential applica-
tion in security.
The safety field also make use of techniques to
predict the performance of systems that are going to
be operated by humans. Predicting how an event can
unfold is highly dependent on the description of the
context, tasks, and failure modes. Potential paths that
an actual event can follow are usually represented in
binary trees called event trees (see THERP for in-
stance (Swain et al., 1980)), where branches represent
what may happen when an event (a leaf) succeeds or
fails. Eventually, probabilities are computed for each
outcome and recommendations are produced to en-
hance the reliability of the system.
This prospective approach is used in Human Re-
liability Analysis. It relies ‘heavily on process ex-
pertise to identify problems and its methods them-
selves or expert estimation for quantification’ (Bor-
ing, 2012). The overall process for a prospective anal-
ysis follows the same processes introduced earlier for
an RCA, that is Figure 2, except for step (ii) which is
called Prospective analysis.
However, there are a few key differences between
the approach in safety and an approach that we devise
in security. They emerge preponderantly and make
migrating the existing retrospective and prospective
techniques from safety to security be not a straightfor-
ward task. Such differences, which bring up several
challenges that need to be addressed and resolved,
emerge in the four steps of the RCA. Table 1 sum-
marizes the differences which raise a five major chal-
lenges: (C
1
) Addressing the lack of knowledge and
structured data: The challenge is to compile and for-
mat factual information about the investigated attack
to allow for the RCA to be performed, to describe
what the attacker does and what are the attacks effects
on the user and on the systems security. Furthermore,
The RCA should provide precise information regard-
ing the data to collect. (C
2
) Investigating Attacks:
The RCA for security must output a set of contrib-
utors and human-related factors that are likely to ex-
plain the success of attacks, or potential attacks. The
new analysis should safeguard against one inherent
shortcoming of RCA: the possible lack of objectiv-
ity. (C
3
) Creating reusable knowledge: To integrate
with existent computer securitys techniques, the RCA
technique should provide direct links between the at-
tackers capabilities and their effects on a systems se-
curity. The challenge is to be able to augment a said
threat model with capabilities that an attacker can gain
by performing user-mediated attacks allowed by the
threat model. (C
4
) Match patterns of known attacks:
The RCA, in addition to the retrospective analysis of
past attacks needs to provide a socio-technical secu-
rity analysis where, from a systems description, so-
cio-technical vulnerabilities, along with their contrib-
utors are listed. (C
5
) Being flexible: The new method
should be flexible enough to adapt to new threat, at-
tacks, and technologies.
5 OUR PROPOSAL: S·CREAM
Figure 3 shows the steps of the process that we pro-
pose to address the five challenges. We describe each
steps separately, hinting where necessary to the tool
that we have implemented to support the execution of
the steps.
From Situation Awareness to Action: An Information Security Management Toolkit for Socio-technical Security Retrospective and
Prospective Analysis
217
Table 1: Key differences in safety and security.
Safety Security
Data Collection &
Investigation
There is an established process to collect
structured evidence for root cause analysis.
This process is not well-established; data are often unstructured
and the information is often scattered across multiple actors.
There are no malicious actors. Incidents hap-
pen because of general malfunctioning.
Incidents are caused by attackers whose skills and capabilities
may be subtle and even unknown.
Accidents to be investigated usually take place
in well-known and well-defined settings
We face much more heterogeneous contexts, furthermore the
incidents can still be unfolding at the time of analysis.
Analysis
RCA techniques are widely used and the hu-
man component is a central part of practices
The use of RCA methods is often advocated but lacks human-
related insights.
A human error is a well defined concept that
can be the starting point of an analysis.
The human error is considered a systems failure mode that does
not call for investigations.
There is always some root cause that can be
isolated for an incident.
The root cause of the success of an attack is always the attacker,
therefore we are interested in all the factors that contribute to
the success of attacks.
The analysis begins from the terminal point of
failure: the observable incident.
An attack/incident can be an intermediate step leading to other
attacks/incidents. Therefore we might not be able to observe
the factual consequences of an attack/incident on a system.
Recommendation
Generation
Removing the root cause prevents the incident
from reoccurring.
Since the root cause is the attacker, technical controls can be
applied on to reduce the attacker’s capabilities. Socio-technical
controls can be applied on the human contributors.
An adverse event, being coincidental, may
never reoccur on similar systems.
An attack incident will re-occur because attackers actively
probe similar systems to recreate it. The sharing of recom-
mendations is thus critical.
Implementation
People involved in incidents are mostly
trained professionals (e.g., pilots, air traffic
controllers, power plant operator).
People are much more diverse with regard to their relevant
skills and knowledge (e.g., children, bank employees, elderly
people, medical doctor). Furthermore they can have motives
and concerns unrelated to security.
Root causes are identified and controlled. It may be impossible to control all identified contributors that
will be actively manipulated by the attacher
Data Collection & Investigation. This step re-
mains, in its goal, as it were in safety. It is about
gathering factual information about an attack, usually
though digital forensics, live monitoring or in-person
investigation. However, to solve challenge C
1
, we
need a description of an attack that is expressed in
an appropriate structured format. This is missing in
the socio-technical security counterpart. We define
an attack description scheme, a structured descrip-
tion of fields as: Attackers actions, their Effects on
the systems security, the Declared and Imitated Iden-
tities, the Command the user is asked to execute, the
Medium used to launch the attack, and the attack’s
Prerequisites. Prerequisites are the capabilities that
an attacker is required to possess in order to perform
the attack. This choice of a structure makes also pos-
sible to generalize the attack’s description, abstracting
from the specific context and system. This is required
if we intend to reuse the knowledge gained about this
attack in the prospective analysis.
Retrospective Analysis. In this steps the analyst
first builds a list of Error Modes, then he searches
for Contributors for them. For this transposition from
safety to security to work, we consider a human er-
ror as being ‘an action or decision that results in
one or more unintended negative outcomes’ (quoted
from (Strauch, 2004)).
An Error Mode (EM) is the analogue of a Fail-
ing Mode in technological failure analysis. An EM
describes a users action that, in judgment of the ana-
lyst’s observation, should have not happened, or not
at that time, or not onto that object.
A Contributor is a characteristic pertaining the
system’s functioning that has facilitated the attacks
success. For instance, ‘Habits and Expectations’ is
a contributor to an e-mail phishing attack if the mali-
cious message is received when a genuine message is
expected (e.g., a monthly email reminding your sub-
scriptions).
We have implemented this step by customizing an
existing technique called CREAM. In so doing we
ICISSP 2017 - 3rd International Conference on Information Systems Security and Privacy
218
address challenges C
2
and C
3
. In CREAMs original
retrospective analysis, the analyst follows a cause-
consequent process which is represented by tables.
By traversing the tables, the analyst searches for an-
tecedents of each Failing Mode. This process is re-
cursive: intermediate ‘generic’ antecedent can be jus-
tified by other antecedents until the analyst finds an-
tecedents that are ‘sufficient in themselves. They are
called ‘specific’ and are the most likely cause of the
inspected incident. In Figure 4, left side, this is the
‘Yes’ branch.
To adapt this process in security, we need to define
a less restrictive stop rule to yield Contributors. We
have to avoid pointing invariably to the attackers ac-
tions and continue to investigate additional contribut-
ing antecedents.
Our method’s stop rule has been redefined in such
way that all likely specific antecedents for the event
are presented to the analyst together with the specific
antecedents that are contained into sibling generic an-
tecedents. Figure 4, right side, shows this process.
The tool implements the process. It shows the an-
alyst with a tree-like structure that the analyst can tra-
verse opening new nodes until he finds a stop condi-
tion (see Figure 5). The analyst uses the description
of the attack to define the security-critical actions car-
ried out by the victim and the associated EMs. Addi-
tional EMs may have to be analysed in the course of
events that lead to the critical action, for instance if
the victim first encounters the attacker and misiden-
tifies him/her as being trustworthy. Considering each
antecedent with the attacks description in hand, the
analyst follows the stop rule to build the list a Con-
tributors of the attack under scrutiny.
Generalization. Generalisation partially addresses
challenges C
4
and C
5
. It is a new and sophisticated
step that requires several successful steps of Data Col-
lection & Investigations and Retrospective Analysis
steps before it can produce valuable outputs. The
output is a list of Attack Modes (AMs) compiled by
grouping the output of several previous steps and or-
ganized into a catalogue. An AM is a link between an
attacker’s capability and the effects that it can produce
on a system’s security. For instance, sending a mes-
sage that nudges a user to click on a malicious link (an
attacker capability) is what allows the attacker to ex-
ecute code on the system (effect produced on the sys-
tem). Once an initial list of AMss is set (and we have
bootstrapped our tools with several of them, see later),
an analyst can use the catalogue to probe, prospec-
tively, a given system for socio-technical vulnerabili-
ties given a threat model (step (iv)).
To implement this step in our tool, we were in the
need to bootstrap the tool’s with an initial catalogue
of AMs. Instead of waiting for a sufficient amount of
real attacks to be observed, described and then anal-
ysed for their root causes, we resorted to look into
fifteen Attack Patterns among those described in the
Common Attack Pattern Enumeration and Classifica-
tion (CAPEC) library (MITRE, 2014). We are here
referring to the library as it were in January 2016.
Then we run step (i) and (ii) on them using the tool.
For reason of space we cannot give more details of
our reasoning as analysts, but we resorted at our ex-
perience of computer security specialists and to com-
mon sense and knowledge in security analysis when
it was necessary to take a decision.
From the CAPEC library we selected the attacks
where the user is at the source of the success of the
attack; after processing them, we populated the AM
catalogue with 29 contributors to two socio-technical
capabilities that we identified. These capabilities are
intermediate goals of the attacker, peripheral and de-
coupled from the system on which they are exploited:
Identify spoofing which is the capability to usurp an
identity, and Action spoofing which is the capability
to deceive the user into thinking that an action he per-
forms will behave as he expects, whereas another ac-
tion, which is harmful for the system, is executed in
its place. Some of these AMs appear in Figure 5.
Prospective Analysis (Security Analysis). The last
step is that of the prospective security analysis. It
also addresses partially challenges C
4
and C
5
. Assum-
ing that a catalogue of AMs has been bootstrapped,
the implementation of this step consists of querying
the catalogue for the Contributors linked to a given
potential threats. The tool implements this step in a
semi-automated manner: the analyst describes the at-
tacker’s capabilities against which he wants to test a
system’s security (i.e., he specifies the threat model)
while the tool filters and displays the corresponding
AMs.
The analyst is supposed to follow a stress-test
driven Security Analysis, as if he were attacking the
system. To ensure that the potential attacks are in-
deed feasible on the system, our tool filters out the
attacks that exceed the attackers capabilities that are
not part of the threat model. From the tool’s view-
point, the attacks that the analyst imagines to launch
on the system are no different from regular attacks
and are therefore investigated in the same way (i.e.,
by steps Data Collection & Investigations then Retro-
spective Analysis). Contributors that links to the ana-
lyst’s attacks are displayed along with the AMs with
the difference that now the AMs now instruct the ana-
lyst about possible ways a similar attacker than him
From Situation Awareness to Action: An Information Security Management Toolkit for Socio-technical Security Retrospective and
Prospective Analysis
219
Figure 4: CREAM process (left) and its adaptation to security (right).
could intrude the system, whereas the Contributors
yielded from the analyst-driven Security Analysis in-
forms the analyst how an attacker could realize the
attacks adverse effect on the system. It is then the
analyst that evaluates whether the system is already
sufficiently protected against the risk of exploitation
of each identified Contributor. If the analyst consider
that the risk is too high, he is left with the duty to pro-
vide controls that reduce the likelihood of a successful
exploitation of this Contributor to an acceptable level.
5.1 The Tool: S·CREAM Assistant
We developed a S·CREAM’s companion tool, the
S·CREAM Assistant which is available at (Huynen,
2016). We wanted the application to be multi-
platform, portable, and stand-alone in the first itera-
tions, while still being able to transpose it to a client-
server model, or even to a desktop application if we
later decide so. Consequently, we chose to implement
S·CREAM Assistant in JavaScript, storing all data in
the web browser’s Local Storage. S·CREAM Assis-
tant uses several frameworks. AngularJS (Google,
2016) manages the Views and the Controllers, js-
data (Js-data Development Team, 2016) handles the
Models and provides an Object-Relation Mapping,
and D3.js (Bostock et al., 2011) displays the interac-
tive tree used to visualise S·CREAM’s Retrospective
Analysis step. S·CREAM Assistant allows the analyst
to customize CREAM’s original tables used by the
S·CREAM methodogy via XLST stylesheets. The an-
alyst can also import and export his analyses stored in
the web browser’s Local Storage into a JSON repre-
sentation.
6 USE CASE
We put ourselves in the shoes of a security practition-
ers who after having assessed risks posed to one of its
organization’s application decides to implement OTPs
to authenticate users on this application. He chooses
the Yubikeys nano USB security token.
A YubiKey is a multi-purpose security token in
the form of a USB dongle. A YubiKey is versatile
as it can present itself as a keyboard or a two-factor
authentication device to a computer (or via NFC to
a smart phone). A YubiKey can be used to generate
and store a 64 characters password, generate OTPs,
or to play different challenge-response protocols (yu-
bico AB, 2015). YubiKey’s user interface consists of
only one button and a LED.
YubiKeys have peculiar user interface and user in-
teractions as they are not doted of a screen. The ab-
sence of a screen has for consequence to shift the duty
of providing feedback to the user to a LED light. The
LED’s behaviors (e.g., flashing rapidly, being on or
off, et cetera) have different meanings and are ex-
plained in the user’s manual (yubico AB, 2015).
One aspect of Yubikeys is that they support two
configuration slots on one device. To use these config-
urations, the user touches the button of the device for
ICISSP 2017 - 3rd International Conference on Information Systems Security and Privacy
220
Figure 5: Screen-shot of the retrospective analysis as per-
formed by using our tool. ‘GA are generic antecedent, ‘SA
specific antecedent. Red antecedent cannot be further ex-
panded, they denote a stop condition in the search for con-
tributors.
different periods of time. These slots can be config-
ured to generate OTPs or a static password. Quoting
from the yubico’s YubiKeys security evaluation docu-
ment (yubico AB, 2012) this functionality assumes a
few security implications:
The YubiKey 2.0 introduces a mechanism
where the user can use two separate creden-
tials. We call the storage for these credentials
‘slot 1’ and ‘slot 2’. To generate a credential
from slot 1, the user touches the button for a
short period of time (e.g., well below 2 sec-
onds). To generate a credential from slot 2,
the user touches the button for a long period of
time (e.g., well above 3 seconds). With proper
user education, we believe this does not add
any additional security problems so we con-
tinue to evaluate the YubiKey configured with
just the slot 1 credential.
It is worth noting that YubiKeys (now in version
4) come in two forms. One is the standard YubiKey,
which is 18 × 45 × 3mm. It has a round-shaped but-
ton with a LED at the top. The other is the YubiKey
‘nano’. Much smaller, it completely disappears into a
USB port when plugged in, and it has button and LED
on its edge.
Without more details about the organization’s con-
text, we only investigate the consequences of config-
uration and functional choices related to the Yubikey
itself. How the different configuration settings can
impact the token’s operations and the provided secu-
rity when used by the organization’s employees?
We perform our analysis on the basic operation of
a YubiKey with the ‘Dual configuration’ functionality
enabled. We set a YubiKey nano to yield an OTP on
slot 1, and a static password on slot 2. The security
practitioner’s rationale being that it would be a waste
not to propose to improve employees’ passwords with
a long random string of characters while the Yubikey
provides this feature.
6.1 Security Analysis
Threat Model. The main assumptions for this sys-
tem are that the attacker can read and write on the
Internet. This Threat Model implies that the attacker
is free to send messages on the web medium to the
user before and after the operation of the Yubikey by
touching its button. More specifically, we consider
that the user is visiting a website under the control of
the attacker.
Semi-automatic Security Analysis. We consider
that this system’s Threat Model allows the attacker to
control the source, the declared identity, the imitated
identity, the command, and that it can write on the
web medium. As the attacker has no control over the
YubiKey, he cannot spoof the action the user is about
to perform. The attacker has control of the sequence
of communication with the user. In consequence, by
using S·CREAM, we find that the reachable Socio-
Technical Capability (STC) is Identity spoofing.
Analyst-driven Security Analysis. To find likely
potential attacks on this system, our strategy is to
formulate hypotheses about the consequences of the
user’s actions in consideration of the attacker’s ex-
tended capabilities.
There are two actions that a user has to carry out
when using a YubiKey on a computer: plugging the
YubiKey into a usb port, and operating the YubiKey
From Situation Awareness to Action: An Information Security Management Toolkit for Socio-technical Security Retrospective and
Prospective Analysis
221
by touching its button according to the authentication
scheme of the application. On the YubiKey nano, both
actions are critical from a security point of view.
plugging the Yubikey nano in a computer can ac-
cidentally produce an OTP because of the location
of the button at the edge of the device. Plugging
or unplugging a YubiKey nano can lead to a loss
of confidentiality of the OTP code located in the
first slot. As the YubiKey operates after the touch-
ing event is finished we consider that the Error
Mode to investigate is ‘Sequence-Wrong action’,
and that the user appends an irrelevant action to
the sequence of actions.
operating the YubiKey nano has two important di-
mensions: the action’s duration (i.e., less than 2
seconds or more than 3 seconds) and the action’s
location (i.e., which user interface element has the
focus at the time of the action). The user needs to
touch the device within the right amount of time
while being in communication with the correct en-
tity; otherwise, there can be a loss of confidential-
ity. As location-based attacks are already covered
by the Identity spoofing (i.e., the user misidenti-
fies the attacker for another entity), we focus on
the duration. In particular, we investigate the EM
‘Duration-Too long’.
Table 2 sums up the results of this investigation.
Discussion on the Results and the Possible Remedi-
ations. Regarding Identity spoofing, the attacker has
a lot of options when it comes to impersonate another
entity (see the Identity spoofings column in Table 2).
A prominent example of such attack is the Man In
the Browser attack: the attacker, in control of the web
browser, redirects the user to a website he controls
when the user attempts to go to his bank’s website.
The attacker then asks for the credentials (including
two-factors Authentication credentials as the one pro-
vided by a YubiKey) and logs into the bank’s website
in place of the user. The key result of this analysis is
that there is little that can be done to thwart the attack,
given the number of Contributors. The results of this
analysis come to the same conclusion as the security
evaluation made by yubico (yubico AB, 2012), which
states, ‘We conclude that the system does not provide
good defence against a real-time man-in-the-middle
or phishing attack.
Regarding potential attacks on the ‘Dual config-
uration’ functionality, Table 2 shows that there are
three Contributors that an attacker can manipulate to
foster the occurrence of the ‘Sequence-Wrong action’
EM during the plugging critical action. The attacker,
in control of the webpage can emit sounds or noises
to apply pressure on the user, and he can also create a
competing task. We see little practical application of
this attack.
Finally, we turn to the case of the operating crit-
ical action. Investigating this critical action with
S·CREAM yielded more Contributors than the plug-
ging critical action, and therefore, it appears more
likely to observe potential attacks that exploit the op-
erating action as opposed to the plugging action. Ta-
ble 2 lists the Contributors that we reckon can be used
to trigger to the ‘Duration-Too long’ EM. For in-
stance, we select ‘SA-Confusing symptoms’ because
the attacker can attempt an attack in the same fashion
as Social-Engineering attacks in which the attacker
sends a ‘bad authentication’ message as sole feedback
after each login attempt, nudging users to give away
every password they know while trying to authenti-
cate. The difference being that, in our use case, the
user would try every possible action on the YubiKey
instead of entering passwords. This kind of attack is
very well possible given the fact that the YubiKey pro-
vides little feedback when a slot is yielded and no
feedback about which slot is yielded. Furthermore,
the user might be unsure how he configured his Yu-
biKey (and someone may have configured it for him).
In the light of this analysis, we consider that the
choice of the Yubikey for implementing OTPs is not a
bad choice, but that the security practitioners should
refrain from using the second slot as it poses some
socio-technical security issues.
7 DISCUSSION & CONCLUSION
We developed S·CREAM with the ambitious aim to
help security practitioners design effectively secure
systems. The toolkit can be used at different points
of the ISM cycle and allows its users to prospectively
(as in the use case presented in § 6) and to retrospec-
tively identify socio-technical factors that could lead
or could have led to a security breach. Used with
discernment these insights can potentially be trans-
formed in requirements or strategies that improve a
system’s security, but one needs to be aware of the
limitations of S·CREAM before using the toolkit.
The first limitation is that our methodology has
to be further developped and properly validated. In-
deed, both the tables inherited from CREAM —the
RCA in safety from which S·CREAM is inspired—
and the catalogue of Attack Modes require care and
maintenance before the methodology can reach its
full potential. Clearly, the catalogue that we boot-
strapped from the CAPEC library is, for the need
of S·CREAM, still rudimentary and has to be re-
ICISSP 2017 - 3rd International Conference on Information Systems Security and Privacy
222
Table 2: Contributors yielded by the Security Analysis.
STC: Identity spoofing Attack: Foster ‘Sequence-Wrong
action’ EM on plugging
Attack: Foster ‘Duration-Too long’
EM on operating
GA-Faulty diagnosis GA: Sound GA: Adverse ambient conditions
GA-Inadequate quality control SA: Competing task SA: Confusing symptoms
GA-Inattention SA: Design SA: Inadequate training
GA-Insufficient knowledge SA: Noise SA: Information overload
GA-Mislabelling SA: Mislearning
GA-Missing information SA: Multiple signals
GA-Wrong reasoning SA: New situation
SA-Ambiguous label SA: Noise
SA-Ambiguous signals SA: Overlook side consequent
SA-Ambiguous symbol set SA: Too short planning horizon
SA-Competing task SA: Trapping error
SA-Erroneous information
SA-Error in mental model
SA-Habit, expectancy
SA-Hidden information
SA-Inadequate training
SA-Incorrect label
SA-Mislearning
SA-Model error
SA-Overlook side consequent
SA-Presentation failure
SA-Too short planning horizon
fined. Furthermore, the antecedent-consequent tables
we use are still not dedicated to security and there-
fore the Contributors yielded by S·CREAM’s Secu-
rity Analysis are currently more generic than what
we think they should be. Because of it sometimes
they are difficult to be interpreted, but we expect the
toolkit’s relevance to grow over time through its use
and with the adjustments that we will make along the
way. Therefore, we consider the current state of the
toolkit as a proof of concept that as future work we
will challenge through the analysis of different secu-
rity incidents and systems, and that we will tune to
improve its accuracy.
Another critical point, is that the analyst is not
necessarily an expert on all aspects upon which
S·CREAM can shed light. This shortcoming is even
more salient when we consider the possible additions
we can make to its tables. For instance, there is a
lot of literature on warnings and how warnings can
have a negative impact on user’s decisions if not im-
plemented properly. If this literature were to make
its way into S·CREAM’s tables with new possible an-
tecedents, the analyst would be expected to be able to
decide whether the warnings that are presented to the
user fulfill their mission.
A final warning is that S·CREAM’s results can be
misused by the analyst. Indeed, it needs to be clear
that the results obtained through the use of the toolkit
are potential not verified causes for an attack. The
toolkit produces a list of potential factors that an at-
tacker may exploit to perform an attack on a system,
but is the analyst’s duty to ponder on whether these
factors should be controlled on the system or not.
7.1 Future Work
We intend to validate the toolkit. By validating
we mean in particular ensuring that the S·CREAM
methodology yields as often as possible sound results
for its security analysis.
While we have assumed the CREAM’s tables also
work for security still we intend to identify the Con-
tributors that are the most often discarded by analysts,
and for this we plan to challenge their relevance ex-
perimentally.
Other improvements concerns the S·CREAM As-
sistant. We intend to design helpers, such as check-
lists, to offer guidance to the analyst who run the tool
and add the possibility to share schemes, catalogues
of AMs, and sets of attacks. The tool works better if
analysts of different companies cooperate and share
their knowledge.
The antecedent-consequent tables inherited from
CREAM should be specialized for security. We in-
tend to provide up-to-date tables of antecedents that
reflect the current state of the research on factors that
influence security-related behavior
However, there is a main obstacle to overcome
before reaching these milestones: we need to find a
sufficient number of documented attacks to analyse.
From Situation Awareness to Action: An Information Security Management Toolkit for Socio-technical Security Retrospective and
Prospective Analysis
223
These attacks and the corresponding attacker’s traces
are in the hand of security practitioners. Therefore,
we need to solve very practical issues in order to make
our toolset usable and used by them before starting
to improve its retrospective and prospective predic-
tion: we have to find a way to access and then to ex-
tract from raw logs the information useful to correlate
socio-technical attacks with user actions and attacker
activities.
We believe that the shortcomings we identified
can be fixed, and that by improving S·CREAM’s ta-
bles, by maintaining our toolset’s knowledge of se-
curity and human-related factors, and by fostering its
use and the sharing of experiences, our toolkit can be
a useful addition to a security practitioner’s toolbox,
but we need to fully implement such ameliorations.
REFERENCES
Adams, A. and Sasse, A. (1999). Users Are Not the Enemy.
Comm. ACM, 42:40–46.
Anderson, R. J. (2008). Security Engineering: A Guide to
Building Dependable Distributed Systems. Wiley.
Beautement, A., Becker, I., Parkin, S., Krol, K., and
Sasse, M. A. (2016). Productive Security: A Scal-
able Methodology for Analysing Employee Security
Behaviours. In Proceedings of the Symposium on Us-
able Privacy and Security (SOUPS) 2016. USENIX
Association: Denver, CO, USA. in press.
Bianco, D. (2014). The pyramid of pain. Avail-
able at http://detect-respond.blogspot.lu/2013/03/the-
pyramid-of-pain.html.
Boring, R. L. (2012). Fifty Years of THERP and Human
Reliability Analysis. Proceedings of PSAM11.
Bostock, M., Ogievetsky, V., and Heer, J. (2011).
D3: Data-driven documents. Available at
http://vis.stanford.edu/papers/d3. IEEE Trans.
Visualization & Comp. Graphics (Proc. InfoVis).
Boyd, J. (1995). The essence of winning and losing.
Brumfield, J. (2015). 2015 Data Breach Investigations Re-
port. Technical report, Verizon.
Caralli, R., Stevens, J., Young, L., and Wilson, W. (2007).
Introducing octave allegro: Improving the information
security risk assessment process. Technical Report
CMU/SEI-2007-TR-012, Software Engineering Insti-
tute, Carnegie Mellon University, Pittsburgh, PA.
Cotroneo, D., Paudice, A., and Pecchia, A. (2016). Au-
tomated root cause identification of security alerts:
Evaluation in a SaaS Cloud. Future Generation Com-
puter Systems, 56:375 – 387.
ENISA (2016). Annual Incident Reports 2015. Technical
Report October, ENISA - European Union Agency for
Network and Information Security.
Ferreira, A., Huynen, J., Koenig, V., and Lenzini, G. (2015).
In Cyber-Space No One Can Hear You S·CREAM -
A Root Cause Analysis for Socio-Technical Security.
In STM, volume 9331 of Lecture Notes in Computer
Science, pages 255–264. Springer.
Google (2016). AngularJS. Available at
https://angularjs.org/.
Huynen, J. (2016). S·CREAM Assistant, a tool
to support S·CREAM analyses. Available at
https://github.com/gallypette/SCREAM-Assistant.
International Organization for Standardization, Geneva, S.
(2005). ISO/IEC 27001:2005 - Information technol-
ogy Security techniques Information security man-
agement systems – Requirements. Technical report.
Ishikawa, K. and Ishikawa, K. (1988). What is Total Quality
Control? the Japanese Way. Prentice Hall.
Js-data Development Team (2016). Js-data. Available at
http://www.js-data.io/.
Kasikci, B., Schubert, B., Pereira, C., Pokam, G., and Can-
dea, G. (2015). Failure sketching: A technique for
automated root cause diagnosis of in-production fail-
ures. In Proceedings of the 25th Symposium on Oper-
ating Systems Principles, SOSP ’15, pages 344–360,
New York, NY, USA. ACM.
Kirlappos, I., Parkin, S., and Sasse, M. A. (2014). Learn-
ing from “shadow security:” why understanding non-
compliant behaviors provides the basis for effective
security. In Proceedings 2014 Workshop on Usable
Security. Internet Society.
MITRE (2014). CAPEC - Common Attack Pat-
tern Enumeration and Classification. Available at
https://capec.mitre.org/.
Noureddine, M., Keefe, K., Sanders, W. H., and Bashir,
M. (2015). Quantitative security metrics with human
in the loop. In Proceedings of the 2015 Symposium
and Bootcamp on the Science of Security, HotSoS ’15,
pages 21:1–21:2, New York, NY, USA. ACM.
Reason, J. (1990). Human Error. Cambridge University
Press.
Schneier, B. (2014). The future of incident response.
Schoenfisch, J., von St
¨
ulpnagel, J., Ortmann, J., Meilicke,
C., and Stuckenschmidt, H. (2015). Using abduc-
tion in markov logic networks for root cause analysis.
CoRR, abs/1511.05719.
Strauch, B. (2004). Investigating Human Error: Incidents,
Accidents, and Complex Systems. Ashgate Pub Ltd.
Swain, A., of Nuclear Regulatory Research, U. N. R. C. O.,
and Guttmann, H. (1980). Handbook of Human Relia-
bility Analysis With Emphasis on Nuclear Power Plant
Applications - Draft Report For Interim Use and Com-
ment. NUREG/CR. U.S. Nuclear Regulatory Com-
mission.
yubico AB (2012). Yubikey security evaluation: Dis-
cussion of security properties and best prac-
tices. Available at https://www.yubico.com/wp-
content/uploads/2012/10/Security-Evaluation-
v2.0.1.pdf.
yubico AB (2015). The yubikey manual: Usage,
configuration and introduction of basic con-
cepts. Available at https://www.yubico.com/wp-
content/uploads/2015/03/YubiKeyManual v3.4.pdf.
ICISSP 2017 - 3rd International Conference on Information Systems Security and Privacy
224