Towards Automated Modelling of Large-scale Cybersecurity
Transformations: Potential Model and Methodology
Artur Rot
a
and Bartosz Blaicke
b
Department of Information Systems, Wroclaw University of Economics, Wroclaw, Poland
Keywords: Security, Cybersecurity, Transformation, Investment, Budget, Risk-based Approach.
Abstract: The purpose of this paper is to propose a proprietary methodology and model to generate a “cybersecurity
transformation workplan” for large organizations that can improve their cybersecurity posture. The key input
is based on risk-based assessment or maturity-based questionnaires depending on existing governance
processes and available information. The original scoring can be then used to prioritize a portfolio of all
possible initiatives by selecting the ones that are missing from typical foundation elements or would have
high potential impact in relation to required investment and effort. Additional constraints such as budget
limitation and FTE availability, logical sequencing and time requirements could be added to ensure effective
use of company resources and actionability of the recommendations. The Gantt-like output would ease the
burden on the security teams by providing an individualized set of activities to be implemented to improve
risk posture.
1 INTRODUCTION
Cybersecurity is a field that has been gaining
significant attention among IT professionals and the
general public. Part of the reason is due to rising
prominence of cyber-attacks and their impact so that
they are now considered one of the top 5 global risks
on par with extreme weather events and natural
disasters (WEF, 2019). However, even considered on
their own, cyberattacks are becoming increasingly
dangerous. They are growing in quantity at a 34% per
year (US-GAO, 2016), sophistication such as recent
Triton attack that could override itself to cover its
tracks (Venkatachary, Prasad and Samikannu, 2018)
and reach as single modern ransomware (e.g.
NotPetya) is able affect computers in more than 100
countries worldwide (Jasper, 2017). Unfortunately,
the defense perimeter is enormous, and attackers only
need a single-entry vector to be successful so not only
many attacks go unnoticed with average time to
detection being multiple months, but many more are
believed to go completely unnoticed (Verizon, 2017).
There is also significant innovation element in
cybersecurity with more than USD 20bn of merger
and acquisition spending in the space and additional
a
https://orcid.org/0000-0002-7281-8253
b
https://orcid.org/0000-0002-5083-0059
USD 5bn being invested by private equity companies
into disruptive cybersecurity start-ups in 178 deals
during 2017 alone (McAlpine et al, 2018).
Regrettably, some of that innovation causes more
confusion with companies having to setup and
manage multiple provider ecosystems each operating
separately and adding complexity that often is not
leading to more security but opening another potential
attack vector for attackers due to misconfigurations.
Internet of Things (IoT) adds another layer of
complexity with the proliferation of over 30bn
connected devices predicted to be online by 2023
(Ericsson, 2017) that can often be easily disoverable
via “Google-like” search engines such as SHODAN
and accessed by anyone using default passwords.
Such services expose hundreds of thousands IoT
devices, many with unchangeble default passwords
and no future firmware updates. (Rot and Blaicke,
2017).
Attacks have also already crossed the digital-
physical barrier as long ago as 2010. We have seen
that cyberattacks can cause physical consequences or
even be used as military weapons that are highly
targeted and potentially more effective than
traditional warfare such as Stuxnet (Zetter, 2014).
Rot, A. and Blaicke, B.
Towards Automated Modelling of Large-scale Cybersecurity Transformations: Potential Model and Methodology.
DOI: 10.5220/0007763703450350
In Proceedings of the 21st International Conference on Enterprise Information Systems (ICEIS 2019), pages 345-350
ISBN: 978-989-758-372-8
Copyright
c
2019 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
345
Malicious actors are now targeting utilities and
energy sector companies, as much as 68% of all such
companies were hit by a cyber-attack in 2016
(Ponemon, 2017).
All the complexity for businesses is further
amplified by shortage of adequate talent, that doesn’t
even come close to match current demand and will
not do so for the foreseeable future (Libicki, Senty
and Pollak, 2014). That often results in inadequate
attention to the issue despite significant threat that
cyber adversaries pose to our increasingly digital
society.
Therefore, the purpose of this paper is to propose
proprietary methodology that could take away some
of the complexity by automating part of the planning
and give organizations a “fighting chance”. Focusing
their activities on more initiatives that would increase
their odds against potential attacks would be the first
step in that direction and we are hoping this paper
could provide that.
2 STATE OF THE ART
APPROACH TO
CYBERSECURITY
TRANSFORMATIONS
Traditionally cybersecurity transformation planning
has been done either by senior cybersecurity function
employees (i.e. CISO or N-1) that often have many
competing and immediate responsibilities or crises to
handle preventing them from devoting all the
necessary time and attention to such long-term
activities. Therefore, more often due to the increased
scrutiny from the board of directors (Rothrock,
Kaplan and Van der Oord, 2017) 3
rd
party specialists
or management consulting companies are brought in
for several weeks or months to deliver such custom-
tailored plans.
The problem is also being discussed within the
academia with various approaches such as
cybersecurity investment supply chain game theory
model (Nagurney, Nagurney and Shukla, 2015),
focusing on how to optimally invest in cybersecurity
controls when organizations are underinvesting
(Panaousis et al, 2014) or how to use economic
incentives for cybersecurity (Vishik, Sheldon and Ott,
2013).
The use of 3rd party vendor typically requires
significant investment that is not always available for
organizations. On the other hand, none of the
academic approaches have been described and
documented yet in such a way that would allow for
easy implementation or as a matter of fact tested in a
real-life scenario. Therefore, we can build on all these
experiences to enhance the proposed model before
testing that as a real-life scenario.
3 PLANNING CYBERSECURITY
TRANSFORMATIONS
In response to the increased complexity and threat,
companies are significantly increasing spending for
cybersecurity. In 2019 analysts estimate that excess
of USD 124bn will be spend on information security
which is a 12.4% increase from last year (Gartner,
2018). However, that increasing spending is not
necessarily correlated with better security for the
companies as there is a vast spectrum of companies
spending above average on security as proportion of
IT spending achieving similar or even lower security
ratings than their peers. In addition, traditional
cybersecurity focus is on controls and processes.
However, there are further layers that need to be
addressed to fully cover the area including
organization aspects (e.g. organization structure) or
governance (e.g. roles and responsibilities).
Historically, spending is focused on technology such
as firewalls, intrusion prevention systems or identity
and access management solutions but not necessarily
with business outcomes in mind or understanding the
impact these solutions would have on the remaining
part of the organization (Choi et al, 2017).
Cybersecurity transformations in such an
environment needs to consider what is already
established and existing within an organization.
Depending on that, it needs to find a balance between
technical and business-related activities that should
be implemented. In addition, a mix of long-term
efforts and quick wins should be established to
protect against current treats quickly but also start
laying down the foundations for protection
mechanisms of tomorrow.
3.1 Establishing the Baseline
The key input needed before considering what needs
to be recommended as part of a cybersecurity
function transformation is a complete understanding
of the current state. That could be gained following
two different approaches, depending on existing
elements at a given organization:
Risk-based approach that is rooted in a more
detailed assessment of threats, vulnerabilities and
controls that results in risk exposures with various
ICEIS 2019 - 21st International Conference on Enterprise Information Systems
346
likelihoods. It requires more effort and is typically
found only in more advanced organizations.
Maturity-based approach has been historically the
preferred method due to easier implementation. It
has certain limitations as it is often self-reported
and carries a potential danger of targeting a certain
maturity level without an established link to actual
improvement of the security posture.
3.1.1 Risk-based Approach
Risk-based approach is a methodology that encourages
organizations to clearly identify, list and assess risks
that they are exposed to. That understanding is used to
focus effort and resources on addressing highest risks
first and deploying adequate protection to other risks.
To follow a risk-based approach the company needs to
create and maintain a list of potential threats, up-to-
date status of vulnerabilities across the environment
and controls that are in place and planned to mitigate
known vulnerabilities that jointly could also be thought
of as likelihood of the risk. All that will inform the
well-established equation (1).
Risk = Threats x Vulnerabilities x Impacts
(1)
That level of understanding is already very helpful for
the security function on its own. In addition, as it
provides detailed understanding of controls and
vulnerabilities it allows to effectively target the
controls and initiatives with the potential to maximize
vulnerability reduction.
3.1.2 Maturity-based Approach
Alternatively, a simpler and still very popular method
is to follow a maturity-based approach. Its goal is to
showcase how advanced any organization is by
assigning a cybersecurity preparedness score or a rank
to each area, often on a 5-point scale from 0 being non-
existent, 1 inadequate, to 4 considered advanced or
state-of-the-art. Such maturity score is mostly based on
self-reported questionnaires covering multiple
elements of a cybersecurity organization from
governance, through existing technical controls to
personnel awareness level (often ~100 questions). The
score allows to prioritize areas requiring immediate
improvement.
4 DEVELOPING A PORTFOLIO
OF INITIATIVES
Following initial phase of assessing the current state of
the organization, next step would include comparing
the individual risks or maturity scores with a
proprietary model of all potential activities that can be
recommended. That will inform which actions are
necessary as well as how they should be prioritized in
a most efficient way given existing constraints.
4.1 Segmentation Framework Options
To cover the entire realm of potential initiatives we
need to establish a framework that will be
comprehensive but also intuitive enough for the
cybersecurity function to include it into their
processes. The most obvious choice for such
frameworks are industry standards such as ISO 27001
and 27002 or NIST Cybersecurity Framework (CSF)
but there are also proprietary approaches such as
Resilience Levers used by McKinsey & Company
that should be compared. A brief overview of primary
areas highlighted by each of the above mentioned
frameworks shows different segmentation and level
of detail available as shown in Table 1.
Table 1: Comparison of key areas for each framework.
ISO 27002
NIST CSF
Resiliency Levers
Information
security policies
Identify
Information assets
and related risks
HR security
Protect
Frontline personnel
Organization of
information
security
Detect
Resilience in
enterprise processes
Asset
management
Respond
Incident response
Access control
Recover
Security integration
into technology
Cryptography
n/a
Differentiated
protection for assets
Physical and
environmental
security
n/a
Deploy active
defenses
Operations
security
n/a
n/a
Communications
security
n/a
n/a
System
acquisition,
development and
maintenance
n/a
n/a
Supplier
relationships
n/a
n/a
Incident
management
n/a
n/a
Business
continuity
management
n/a
n/a
Compliance
n/a
n/a
Towards Automated Modelling of Large-scale Cybersecurity Transformations: Potential Model and Methodology
347
Despite the differences, all the activities could
also be mapped against all the frameworks
simultaneously at deeper levels if needed, providing
another layer of comparison.
4.2 Prioritization Approach
Having the framework selected, we would need to
establish a large portfolio of initiatives that would
cover multiple possible scenarios that a company
might find itself in. Sample set of initiatives could
include the following:
Launch phishing campaigns with real-time
feedback for users;
Set common vision and mission for the
cybersecurity function in line with business
objectives;
Introduce Network Access Control (NAC)
solution(s).
The model would capture the following
proprietary attributes for each of the activities in the
baseline portfolio:
Actionable description;
Cybersecurity areas that an activity would
impact based on selected taxonomy;
Estimate of time needed to implement
(adjustable based on company data such as
revenue and FTEs detailed in 5.1);
Sequencing/prioritization within each
taxonomy section (based on logical flow of
implementation activities) and globally (based
on overall impact and interdependencies);
Two high level indicators whether this would
be a “quick win” (less than a month) or a
longer-term endeavor as well as if that would
be a “basic, intermediate or advanced” type of
control once implemented;
Estimated impact on the initial self-assessment
score (maturity or risk-based) if that activity
would be completed;
Flags for omitting an activity (already in place)
or indicating current work in progress or
manual prioritization modifier.
Combination of the input scoring and full
portfolio prioritization would result in each activity
being marked as relevant or not and prioritized
accordingly to their risk reduction or maturity score
improvement to form an individual output. That
allows to accommodate for the fact that not every
element needs to reach the highest possible risk
reduction potential or maturity, nor would it be
effective use of company resources in most cases.
5 GENERATING AND
ADJUSTING THE OUTPUT
PLAN
The final element once the input and initial
prioritization are complete, is to adjust the output and
recommendations in a way that would make it
relevant and usable for an organization. One way to
do it is to allow for certain constraints to be defined
so that the desired plan is relatable and realistic. For
example, if we were to propose an estimated effort to
be 10 times larger than what the organization could
typically handle than such recommendation is not
going to be taken seriously. However, if we can distil
which actions would be the best candidates within the
budget constraints that would result in a solid
foundation that could be used as a starting point by
the security function. Secondly, the output
presentation element is equally important since “a
picture is worth a thousand words” and having a
visual representation with easily accessible details is
crucial.
5.1 Constraints Handling
Each company at any given moment in time is
operating in a unique environment with different
constraints following a different vision and plan. To
be able to make the recommendations relatable, the
activities need to be adjustable based on the following
criteria:
Company size based on revenue or Full Time
Equivalent (FTE);
Allocated cybersecurity budget;
Size of the cybersecurity team;
Number of potential concurrent efforts that
could be handled simultaneously;
Aspiration level to be achieved (e.g. reduction
of risk by X% or maturity score improvement
by Y points).
Modifying these criteria would allow to adjust the
baseline output as close to the desired state as
possible. However, while it is possible to make these
modifications, it is not required as it might also be
beneficial to understand the long tail of what needs to
be done as the first iteration before proceeding to
adjust the scope.
ICEIS 2019 - 21st International Conference on Enterprise Information Systems
348
Figure 1: Potential visual representation of the output.
5.2 Output Generation
The output from the model would be a Gannt chart
(see Figure 1 as an example based on previous
manually prepared case study) offering an overview
of all the recommended activities, put in a sequence
with an estimate of man-hours needed to complete the
entire plan. Such output would be modifiable using
previously discussed constraints criteria that the
organization would like to apply. Proposed
automatization approach should be a ranking or
points-based system with multiple parameters to
initially address the sequencing aspect. In addition, it
should be aided with algorithms for multi-project
scheduling such as presented for IT development
efforts (Chen et al, 2017). The added complexity is
that it would need to operate under time dependent
uncertainty because early in the planning exact
initiatives durations are highly uncertain as proposed
more generally in (Song et al, 2019). Potentially with
large enough sample set a machine learning algorithm
could be used to further automate the process and
improve the output quality. However, early on
following a simultaneous manual process to validate
and adjust the automated output would be key to
ensure high quality and allow for adjustments to the
model.
Such proposed work plan could then be used
directly as a high-level blueprint for an organization
to start improving their cybersecurity function or
further modified by the security team for their
specific needs. Additional adjustments could be based
on state of individual projects or activities that might
overlap with what IT or other adjacent functions are
performing for additional customization.
6 CONCLUSIONS
We began this paper with an overview of how
cybersecurity is gaining importance but remains a
complex challenge for most organizations. After
examining current state, we concluded that such
fundamental issue as holistic improvement of
individual company cybersecurity posture is not
easily solved with currently available methods. Some
of which are too abstract and theoretical while others
do not scale well and require additional expenditure
that might place them out of reach for some
organizations.
Thus, we hope that our early thinking on the
proposed proprietary approach and methodology
described in this position paper provides an
Towards Automated Modelling of Large-scale Cybersecurity Transformations: Potential Model and Methodology
349
explanation of how automating the modelling of
large-scale cybersecurity transformations can be done
in an approachable way. We believe that is an
important problem that we could solve or at the very
least simplify for all the organizations that are looking
to improve their cybersecurity posture in an efficient
way.
Fully developing the described methodology and
model would allow to shorten the time and effort
needed to create such comprehensive transformation
plans and, in some cases, might be enough to get the
company on the right track immediately.
REFERENCES
Chen, R., Liang, C., Gu, D., Leung, J., 2017. A multi-
objective model for multi-project scheduling and multi-
skilled staff assignment for IT product development
considering competency evolution, International
Journal of Production Research, Issue 21, Volume 55,
May 2017.
Choi, J., Kaplan, J., Krishnamurthy, C., Lung, H., 2017. Hit
or myth? Understanding the true costs and impact of
cybersecurity programs. https://www.mckinsey.com/
business-functions/digital-mckinsey/our-insights/hit-
or-myth-understanding-the-true-costs-and-impact-of-
cybersecurity-programs. Accessed Jan 2019.
Ericsson, 2017. Ericsson Mobility Report with Middle East
and Africa Appendix, November 2017.
https://www.ericsson.com/assets/local/mobility-
report/documents/2017/ericsson-mobility-report-
november-2017-middle-east-and-africa.pdf.
Stockholm, Sweden. Accessed Jan 2019.
Gartner Inc., 2018. Gartner Forecasts Worldwide
Information Security Spending to Exceed $124 Billion
in 2019, United States. https://www.gartner.com/en/
newsroom/press-releases/2018-08-15-gartner-
forecasts-worldwide-information-security-spending-
to-exceed-124-billion-in-2019. Accessed Jan 2019.
Jasper, S., 2017. Russia and Ransomware: Stop the Act, Not
the Actor. The National Interest, November 20, 2017.
USA.
Libicki, M., Senty, D., Pollak, J., 2014. Hackers Wanted:
An examination of the cybersecurity labor market,
Rand National Security Research Division. Santa
Monica, California, USA.
McAlpine, E., Tedesco, M., Skirbe, K., Boukouris, D.,
Isagon, J., Keswani, J., 2018. Cybersecurity Almanac.
https://momentumcyber.com/docs/Yearly/2018_Cyber
security_Almanac.pdf. Accessed Jan 2019. Momentum
Cyber, San Francisco, USA.
Nagurney, A., Nagurney, L.S., Shukla, S., 2015. A Supply
Chain Game Theory Framework for Cybersecurity
Investments Under Network Vulnerability. Amherst,
USA.
Panaousis, E., Fielder, A., Malacaria, P., Hankin, C.,
Smeraldi, F., 2014. Cybersecurity Games and
Investments: A Decision Support Approach. Queen
Mary University of London, UK.
Ponemon Institute, 2017. The state of cybersecurity in the
oil & gas industry, United States. http://
news.usa.siemens.biz/sites/siemensusa.newshq.busines
swire.com/files/press_release/additional/Cyber_readin
ess_in_Oil__Gas_Final_4.pdf. Accessed Jan 2019.
Rot, A., Blaicke, B., 2017. Internet of Things security.
Selected threats and protection methods on the example
of manufacturing systems, Publishing House of
Czestochowa Technical University, Czestochowa,
Poland.
Rothrock, R.A., Kaplan, J., Van der Oord, F., 2017. The
Board’s Role in Managing Cybersecurity Risks, MIT
Sloan Management Review Magazine: Winter 2018
Issue. Cambridge, USA.
Song, W., Kang, D., Zhang, J., Cao, Z., Xi, H., 2019. A
Sampling Approach for Proactive Project Scheduling
under Generalized Time-dependent Workability
Uncertainty, Journal of Artificial Intelligence Research
Volume 64, 2019.
United States Government Accountability Office (US-
GAO), 2016. Information Security. Agencies Need to
Improve Controls over Selected High-Impact Systems.
https://www.gao.gov/assets/680/677293.pdf
Washington DC., USA. Accessed Jan 2019.
Venkatachary, S.K., Prasad, J., Samikannu, R., 2018.
Cybersecurity and cyber terrorism - in energy sector
a review. Journal of Cyber Security Technology
Volume 2, 2018.
Verizon, 2017. Data Breach Investigations Report 10
th
Edition, USA.
http://www.verizonenterprise.com/verizon-insights-
lab/dbir/tool/. Accessed Jan 2019.
Vishik, C., Sheldon, F., Ott, D., 2013. Economic Incentives
for Cybersecurity: Using Economics to Design
Technologies Ready for Deployment.
World Economic Forum (WEF), 2019. The Global Risks
Report 2019, 14th Edition. Geneva, Switzerland.
http://www3.weforum.org/docs/WEF_Global_Risks_
Report_2019.pdf. Accessed Jan 2019.
Zetter, K., 2014. Countdown to zero day. Stuxnet and the
launch of the world's first digital weapon, Random
House. USA.
ICEIS 2019 - 21st International Conference on Enterprise Information Systems
350