WILL THE CLOUD MAKE THE CITIZEN MORE
VULNERABLE?
Risk and Vulnerability Assessment in Times of Cloud-computing
Stefan Scheer
1
, Ioannis Kounelis
1,2
and Jan Löschner
1
1
European Commission – Joint Research Centre, Institute for the Protection and Security of the Citizen, Ispra, Italy
2
Royal Institute of Technology (KTH), Stockholm, Sweden
Keywords: Digital Citizen, Digital Society, Cloud Computing, Risk, Hazard, Vulnerability, Exposure.
Abstract: Through digital traces that are left behind citizens are more and more exposing their personal data, digital
identities to third parties in a conscious or mostly unconscious way. The latter is particularly the case for a
variety of cloud computing applications implicitly used by a default citizen. To interact in a digital world
and to give away personalized information opens the door for several hazards that may occur – deliberately
or not. Consequently a citizen becomes vulnerable in various dimensions. Current research tries to re-apply
well-known risk analysis strategies within the new context and in particular to cloud-computing scenarios.
Final aim would be to identify individual risks in a qualitative and quantitative manner.
1 INTRODUCTION
In a growing global digital world a citizen is
nowadays navigating in various dimensions: advan-
ced use of mobile phones, citizen in motion, being
listed in digital archives, putting personalized
information on public platforms, etc. While naviga-
ting, the citizen is creating a steadily increasing trail
of more or less personal data projected and archived
in the digital world.
Cloud applications are more and more becoming
state-of-the-art from various points of view when it
comes to outsource any type of (software-related)
service or the externalisation of data repositories.
Moreover the externalisation of data is done in a ra-
ther hidden way, without explicitly “telling” the user
where his/her data are residing. However, externali-
sation of data transfer and data storage can make a
citizen vulnerable as the handing over of rather per-
sonal or mission-critical data is similar to handing
over your personal belongings to an untrusted per-
son.
Hence a citizen is exposing himself (through his
data) and thus making himself vulnerable and put-
ting himself, his personality, his digital assets and
commitments at risk.
Obviously, at that point it would be interesting to
know up to which extent the classical theory of risk
analysis (and even the use of risk analysis tools),
mainly applied to vulnerabilities from natural ha-
zards, accidents, etc., could be transferred and re-
applied within the context of citizens’ digital inter-
actions and in particular to cloud-based applications.
In addition, as it seems that many security-related is-
sues are tackled as well, it is the intention of current
research to put a focus on the citizen, his vulnera-
bilities, exposure and ways to handle these prob-
lems.
2 CLASSICAL RISK ANALYSIS
Classical risk definitions ((UNDP, 2004) for
example) refer to risk as the combination of the
probability of an event and its negative
consequences. Many refer to a hazard, for the first
part, and to vulnerability, describing the second part.
Hence, risk is a function of two components ‘hazard’
and ‘vulnerability’ whereas the latter frequently is
extended towards ‘exposure’ thus getting the
equations (1) and (2) as follows:
R = H * V (1)
R = H * V * E (2)
Risk (R) is the magnitude of an interaction of a
given hazard (H) with a citizen’s vulnerability (V)
173
Scheer S., Kounelis I. and Loeschner J..
WILL THE CLOUD MAKE THE CITIZEN MORE VULNERABLE? - Risk and Vulnerability Assessment in Times of Cloud-computing.
DOI: 10.5220/0003960801730176
In Proceedings of the 2nd International Conference on Cloud Computing and Services Science (CLOSER-2012), pages 173-176
ISBN: 978-989-8565-05-1
Copyright
c
2012 SCITEPRESS (Science and Technology Publications, Lda.)
and a citizen’s exposure (E) towards this. Equation
(2) has typically been described by (Crichton, 1999)
for the field of risks from natural hazards; Figure 1
shows the dependencies in a graphical way.
Figure 1: The Risk Triangle (Crichton, 1999).
The interpretation of risk means that it is the
probability of harmful consequences or expected
losses resulting from an exposure to a given hazard.
2.1 Hazard
In the context of a digital world, and in particular
while using cloud-computing services, we consider a
hazard as signifying a potentially damaging event,
phenomenon and/or human activity, which may
cause property damage, social or economic disrupt-
ion. Usually hazards are single events; but we can
learn from classical risk analysis that under certain
circumstances risks can also be sequential or combi-
ned and as such can be summed up (UNDP, 2004).
Generally, a hazard is a situation that poses a level
of threat; as such a hazard can be dormant or only
theoretical. However, within a certain probability a
hazard may develop within an incident.
Inherent part of the hazard definition is also the
probability that a particular hazard may evolve. In
that sense, the occurrence of a specific hazard is out
of the scope a citizen is actively navigating in.
The origins or sources for incidents can be mani-
fold: there can be malicious attacks as well as hidden
built-in functions on (digital) applications or profi-
ling agents that combine various sources of informa-
tion in order to derive new data or information
(Wright, 2010).
2.2 Vulnerability
Vulnerability which – in the classical definition – is
very much related to the general living condi-tions
of an individual may – within a wider scope – be
extended away from a pure economical condition
towards a vulnerability scope affecting more and
more the social status of an individual.
With that we mean in particular the affectation
of an individual’s digital identities, devices and
presences on various digital platforms; it can also in-
clude (correct) access to information as well as the
integrity of an individual’s personal data.
Vulnerabilities may arise from situations that can
be questioned like: are my data transmitted in a
secure way? Are my data which are stored in the
cloud available on request? Can my cloud-stored
data be restored totally and correctly? Are my cloud-
stored data only revealed to me and to nobody else?
Do the cloud-based services deliver the right results?
Though not further discussed within the context
of this paper, quantifying an individual’s vulnera-
bility may also need to take into consideration the
individual’s own resilience and ability towards
applying appropriate mitigation measures thus
possibly reducing an individual’s vulnerability.
2.3 Citizen Exposure to Cloud
Applications
Exposure is the third decisive factor in equation (2);
with regard to section ‘2.2 Vulnerability’, an
exposure may establish a situation as described in
that section. Contrarily to exposure to natural
hazards, a citizen’s exposure in the cloud is of less
physical nature rather than of virtual nature. A
citizen may of course exhibit his digital devices to
the world, open up unsecured network connections;
more exposure, however, may be demonstrated by
exhibiting individual information or private data by
using digital applications.
Concerning citizens’ based cloud-computing the
following main types of exposure may be treated:
1. exchange of information and data through
network channels;
2. deposit of personal information and data on
external devices;
3. request of software services and delivery of
results;
4. writing/publishing of personal information/data
on (semi-)public internet platforms.
With the availability and the use of several plat-
forms and/or services, a citizen’s total exposure will
be quantified by summing up every single exposure.
3 CLOUD-DEPENDENT RISKS
3.1 General
Cloud-computing offers a wide range of functio-
nalities and characteristics each of which may be at
stake under certain conditions (Jeffery and
Neidecker-Lutz, 2010). Often cloud-computing
services are used by the citizens without explicitly
CLOSER2012-2ndInternationalConferenceonCloudComputingandServicesScience
174
knowing about it or explicitly asking for it which
means that a default citizen may often not be aware
of the under-lying risks.
From a citizen perspective cloud services are
often requested for storage and for data/information
exchange purposes. Hence PaaS-type cloud services
may be prevalent. Moreover cloud services like
SaaS may be requested as well, while we assume
that from a citizen’s perspective the IaaS-use case
would be the least requested. In addition, our inves-
tigations are focussing on the use of public clouds
rather than private ones.
3.2 Availability
Being one of the core aspects of cloud compu-ting,
the instant availability of cloud services is essential,
simply because each individual strongly relies on the
(implicit) promise to get his/her data back. Recent
cases have shown that unfortunately this is not
always the case (Parnell, 2012a; Parnell, 2012b).
Availability mostly also concerns the instant
back-delivery to the user upon request. Cloud-based
services, however, operate on the basis of resource-
sharing; an instantaneous answer viewable on a
user’s device may be hampered by too many concur-
rent requests.
3.3 Reliability and Quality of Service
The reliable delivery and correct restoration of
outsourced citizen data is crucial for the acceptance
of such services. In that sense the prevention of data
losses as well as the restitution of data in its original
context is of highest priority. Basically cloud
providers are external organisations and as such their
functioning may stop without the individual’s
influence. In the worst scenario the total loss of data
may occur (McKendrick, 2011).
Many use cases also require a high quality of
service thus reducing response times or increasing
the general throughput of data. User vulnerability
may vary according to specific user resilience, for
example to dispense with a timely restoration of own
data.
3.4 Economical and Social Aspects
from a User’s Perspective
Though business-related economical aspects may
play a less critical role for most of the private
citizens’ use cases, it can nevertheless become an
important point in a citizen’s portfolio if economical
aspects are tackled in an unpleasant way. Users are
mostly attracted by free or low-cost services offered
through cloud applications. With other words, if this
aspect may not be valid any more or if the return-on-
investment is too low, a user may easily better
refrain from such services rather than continue to
pay, at the end, even more than originally planned.
An even more interesting aspect could be
derived from an indirect affectation of a citizen’s
assets (financial, health, reputation, etc.). Obviously,
risks to losses or damages of that kind can be tre-
mendous for an individual. Except for the handling
of financial losses, the quantification of social
damage opens up a new field of investigation.
Citizens as private persons may have little
mitigation strategies compared with commercial
companies and are often at the mercy of the
particular situation; in addition, a wide-spread non-
awareness of such risks among the population as
well as lack of relevant knowledge makes it more
and more important for a citizen to assess his/her
individual situation and vulnerability, respectively.
3.5 Psychological Acceptability of
Security
Closely related to non-awareness or lack of relevant
knowledge is the known fact that individuals tend to
underestimate the potential hazard while exposing
themselves too much. More-over general
observations reveal a contradictory be-haviour
towards increased security measures: the installation
of relevant measures does not always demonstrate
the intended results. Users tend to feel
uncomfortable with such measures (Bishop, 2003).
Risks deriving from these considerations will have
to be addressed as well.
3.6 Ease of Use of Cloud Services and
Risks of Negligence
For the individual most of the (hidden) cloud-
computing applications are easy to use; in addition,
several types of (competing) applications attract the
user. Consequently individuals tend to spread their
data (in parallel) on various platforms or with seve-
ral services. The spreading of (same or similar) data
on more platforms, however, contributes to an in-
crease of exposure and thus to an increase of risk.
In addition, there is the fact that one and the
same cloud provider aggregates its services thus
mixing and copying a user’s data with other app-
lications; for example, all Google owned services
bear the risk that a user’s data (and setting) that
originally had been provided for one service, may
become “available” for other services, too. Hence an
increase in a user’s exposure is observed.
WILLTHECLOUDMAKETHECITIZENMOREVULNERABLE?-RiskandVulnerabilityAssessmentinTimesof
Cloud-computing
175
Connected with a user’s easy-to-use attitude is a
general negligence (or sometimes impossibility) to
wipe out personal information or data. In most cases
the original data pertain “for ever”, and mostly also
without the explicit consent or knowledge of the
owner of the data. Recent observations from Social
Media sites have revealed that deletion of user data
does not entirely mean deletion from the server.
3.7 Security and Privacy Concerns
The full range of security and privacy issues apply
for most of the cloud applications, mostly arising
from multi-tenancy and concurrency issues. Data in
the cloud are shared between multiple tenants. The
same resources, e.g. a database, may be
(unknowingly) shared among several users. It is
obvious that risks regarding security and privacy
concerns will arise (Jeffery and Neidecker-Lutz,
2010). What if we envisage a software failure during
data restoration (from a commonly shared)
database?
Moreover the location of data is potentially not
known. For legal purposes it could be extremely
difficult to identify a country-specific legal basis in
case of a dispute.
3.8 Exploitation of Data
Known business models of cloud-service providers
or cloud-computing applications revealed the
massive use of information and data provided by
individuals for the purpose of commercialisation.
Multi-facetted user-provided data can be re-shuffled
or aggregated in many different ways thus creating
assets of interesting commercial value.
Data are “handed over” to external stakeholders
who, in principle, can read, copy, and exploit them
commercially. Though data protection principles
would usually apply within cloud computing
contracts, there is little control over potential ex-
ploitation of citizen-own data. Modern social media
platform even are deliberately exhibiting user data to
more or less defined online communities.
Risks deriving from this way of exploiting user
data are of a new dimension and will have to be
thoroughly assessed and discussed with ongoing
research.
4 DISCUSSION
The aim of the current research is to look into the
qualitative and quantitative dimensions of citi-zens’
vulnerabilities while interacting in the digital world,
and in particular to cloud-computing appli-cations.
Therefore, this position paper investigates on the
application of classical risk analysis concepts to the
particular field of citizen’s digital interaction.
Research will define what the single constituents
of the classical risk formula (hazard, vulnerability,
exposure) mean in the new context; current research
should also give hints on how these concepts will be
classified and quantified. In a further step the risk
formula concept will be applied to a number of com-
mon scenarios out of the cloud-computing world
which can affect the citizen’s assets in general.
Ongoing research will postulate the necessity of
detailed classification and subsequent quantification
of all risk affecting factors or parameters: incidents
for hazard quantification; types of vulnerabilities
including individual resilience; types of exposure
and possible degrees of exposure.
In a different step, it is planned to apply known
risk analysis methods, for example FMEA (Failure
Mode and Effects Analysis) or FTA (Fault Tree
Analysis); eventually results will be produced that
would help to generate a clear statement on the
individual risk a citizen takes while navigating in the
cloud.
REFERENCES
Crichton, D.: The Risk Triangle, in: J.Ingleton ed., Natural
Disaster Management, London: Tudor Rose, pp.102-
103 (1999).
United Nations Development Programme (UNDP):
Reducing Disaster Risk. A Challenge for
Development. A Global Report: UNDP – Bureau for
Crisis Prevention and Recovery (BRCP), New York,
2004. Available at: http://www.undp.org/bcpr/
disred/rdr.htm .
Wright, D., Safeguards in a World of Ambient
Intelligence, Chapter 4: Threats and Vulnerabilities.
Springer, 2010.
Jeffery, K., Neidecker-Lutz, B. (eds.): The Future of
Cloud Computing, Opportunities for European cloud
computing beyond 2010. Expert Group Report, public
version 1.0. European Commission, DG Information
society and Media, 2010.
Parnell, B.A., Millions face Megaupload data deletion by
Thursday, in: <http://www.theregister.co.uk/2012/01/
30/megaupload_users_to_lose_data/>, 2012. Accessed
6/3/2012.
Parnell, B.A., Microsoft’s Azure cloud down and out for 8
hours, in< http://www.theregister.co.uk/2012/02/29/
windows_azure_outage/>, 2012. Accessed 6/3/2012.
McKendrick, J.: Eight cloud computing risks, and how to
quash them, in: <http://www.zdnet.com/blog/service-
oriented/eight-cloud-computing-risks-and-how-to-
quash-them/7752>, 2011. Accessed 5/3/2012.
Bishop, Matt, “Computer Security: Art and Science”.
Boston, MA: Addison-Wesley, 2003, pp. 348-349.
CLOSER2012-2ndInternationalConferenceonCloudComputingandServicesScience
176