Citizen Empowerment by a Technical Approach for Privacy
Enforcement
Sascha Alpers
1
, Stefanie Betz
2
, Andreas Fritsch
2
, Andreas Oberweis
1, 2
, Gunther Schiefer
2
and
Manuela Wagner
2
1
FZI Forschungszentrum Informatik, Karlsruhe, Germany
2
Karlsruher Institut für Technologie (KIT), Karlsruhe, Germany
Keywords: Data Protection, Mobile Apps, Privacy Enhancing Technologies, Data Sovereignty, Legal Conformity,
Usability.
Abstract: It is a fundamental right of every natural person to control which personal information is collected, stored
and processed by whom, for what purposes and how long. In fact, many (cloud based) services can only be
used if the user allows them broad data collection and analysis. Often, users can only decide to either give
their data or not to participate in communities. The refusal to provide personal data results in significant
drawbacks for social interaction. That is why we believe that there is a need for tools to control one's own
data in an easy and effective way as protection against economic interest of global companies and their
cloud computing systems (as data collector from apps, mobiles and services). Especially, as nowadays
everybody is permanently online using different services and devices, users are often lacking the means to
effectively control the access to their private data. Therefore, we present an approach to manage and
distribute privacy settings: PRIVACY-AVARE is intended to enable users to centrally determine their data
protection preferences and to apply them on different devices. Thus, users gain control over their data when
using cloud based services. In this paper, we present the main idea of PRIVACY-AVARE.
1 INTRODUCTION
Our life is characterized by connected services and
ubiquitous internet. Businesses are connected via the
Cloud, citizens use available services via the global
network. This leads to fundamental changes in
society and business life as well as communication
in general. Internet of things, digital social networks,
commercial rebate systems, cloud applications and
ubiquitous services lead to an increasing value of
personal information. This information is collected,
stored, used and exploited (partly without the user
being aware of it) although European data protection
provisions require an adequate level of transparency
(Christl and Spiekermann, 2016, p. 121).
Currently, many services require more privileges
than they actually need. It is a legal obligation to
minimize data collection, but this is not actual
practice (Felt et al., 2011a, 2011b). Moreover, it is
not always necessary, that users are clearly
identifiable via IDs (for instance, in case of iOS the
Unique Device Identifier), but many applications use
this information and transfer it only partially
encrypted to the respective provider (Smith, 2010).
More problems occur, if users disclose also third
persons' personal information (like photos or contact
data) without their consent. This can lead to data
protection infringement (Local Court Bad Hersfeld
15.05.2017 - F 120/17 EASO, 2017).
This situation shows the problems for users to
protect their privacy, especially on different devices.
Moreover, researchers face the so called privacy
paradox: On the one hand users and industry often
express their concerns on phenomena like Big Data
or Internet of Things and the desire for enhancing
privacy. But, on the other hand they use privacy
infringing services without applying privacy
protecting solutions (Vervier et al., 2017). There
exist several approaches to explain this paradox, like
for example the users' missing awareness of privacy
risks due to a lack of proper information (Leibenger
et al., 2016). Or the possible low availability of
privacy preserving solutions and the high effort or
expenses involved (Forum Privatheit, 2014). Also
discussed in research are network effects that
determine which network is used. Think for example
about more privacy sensitive alternatives to
Alpers, S., Betz, S., Fritsch, A., Oberweis, A., Schiefer, G. and Wagner, M.
Citizen Empowerment by a Technical Approach for Privacy Enforcement.
DOI: 10.5220/0006789805890595
In Proceedings of the 8th International Conference on Cloud Computing and Services Science (CLOSER 2018), pages 589-595
ISBN: 978-989-758-295-0
Copyright
c
2019 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
589
WhatsApp like e.g. Threema. They fail to get a big
market share as they enter the market after the big
player (Schreiner and Hess, 2015). Taking this
example, users can decide to either give their data to
the service provider or not to participate in
communities. So it is possible that the refusal to
provide personal data results in drawbacks
participating in digitalized social interaction.
Therefore, it is important to help users to protect
their personal data through user-friendly privacy
enhancing technologies, while still allowing them to
benefit from the full range of useful services often
only offered for free in “exchange” for personal
data.
In Europe, the upcoming General Data
Protection Regulation (GDPR) empowers the
citizens' self-determination and obliges providers to
apply privacy by design. These obligations also
apply to international companies outside the EU, if
they offer goods and services to people located
within the EU or monitor their behaviour. If
providers violate these obligations, technical control
and enforcement mechanisms can empower users to
claim their right to data protection. But, they need to
be legally compliant because if users risk legal
consequences they might feel discouraged from
using privacy solutions. Moreover, the technology
should be user friendly and users should be able to
centrally define privacy settings in accordance to
their individual preferences once, which afterwards
are distributed and enforced on (all) devices of the
user. Thus, in this position paper we are presenting
such an approach: PRIVACY-AVARE
The outline of the paper is as follows: We first
present existing approaches for privacy
enforcements (in section two), then we present our
approach: PRIVACY-AVARE in section three (The
Idea). Finally, we end the paper with a short
summary and outlook.
2 DATA PRIVACY
PREFERENCES
It is already possible to allow or deny access to
certain categories of data on many systems. For
example, since Android 6 it is possible to allow or
deny apps the access to address book, calendar,
location and sensors. But it is not possible to set fine
granular preferences. In a typical address book for
each contact forename, surname, birthdate, email
address, postal address, telephone numbers, mobile
numbers, company information and maybe role
information are stored. For a network application
like WhatsApp or Threema only a few of this
information is needed, typical users need name and
mobile number (and for Threema also mail address).
So users need the possibility to allow apps to access
only certain fields of each address. Additionally, the
users want to hide selected contact information for
example, if the relationship is socially taboo.
Fine granular preferences also exist for other
data categories, e.g. for the location information.
There are some apps / services like turn-to-turn-
navigation that need the exact location. But, for
other apps it is enough to know the position in a
wider radius. For a weather forecast app for example
a lower accuracy would be enough.
We miss this functionality in existing solutions.
Three kinds of approaches do exist from a technical
viewpoint:
1. Remove authorizations via the modification
of the manifest file (e.g. Advanced Permission
Manager)
2. Add a security library (e.g. SRT AppGuard)
3. Make a modification at the operating system
level (e.g. XPrivacy)
Figure 1 shows these three approaches, visualizing
modifications of the (source) code in dark colour.
The latter two solutions use sandboxes. A sandbox is
referred to as environment, which restricts actions
by an application according to defined rules (Bishop,
2012). By an access restriction the risk of a violation
of the defined rules is reduced (Goldberg et al.,
1996). This concept, derived from IT security, was
adopted to data protection, e.g. by Backes et al.
(2015).
We have published an overview of our analysis
on existing solutions regarding access management
for user support (Alpers et al., 2017a), including a
comprehensive assessment concerning usability and
functionality aspects. In the following, we shortly
present the main limitations of the examined existing
solutions:
a) Only few solutions enable the user to reveal
just selected parts of information (e.g.,
chosen contacts (from all) or only telephone
numbers hiding further information).
b) Currently, only a few solutions grant the
option to provide substitute data to ensure
the ongoing performance of the used
services. Existing solutions are for example
PDroid or MoboClean (partially).
c) Users are not supported when applying
privacy settings. For example, very few
solutions provide group settings.
CLOSER 2018 - 8th International Conference on Cloud Computing and Services Science
590
d) Overall, basic usability should be provided,
however it is often overlooked in existing
privacy solutions.
e) Users need to have a relatively high level of
technical proficiency to use the existing
solutions, especially when using those with
more functionalities than just blocking the
data access.
f) Legal compliance is often not ensured for
existing solutions (see 3.3).
Thus, a new, more comprehensive solution for
privacy enforcement is needed.
3 THE IDEA
To empower users to decide themselves about their
own data we suggest more privacy settings options.
A typical user utilizes different apps (like
WhatsApp, Threema, Maps, etc.) and services
(routing, mail, internet, etc.) on different operating
systems (Android, iOS, Windows, Linux) and
devices (smartphones, tablets, notebooks, SmartTVs,
cars, etc.). Many of these devices, systems, services
and apps have access to stored personal information
or the chance to collect information by sensors or
user tracking. For every app and service, a user has
an idea which data may be shared or collected. Even
if every system would allow the user to control the
access, he will not describe the preference manually
on each system for each service or app. It should be
possible to describe them once (for a category of
apps and services) and enforce them on every
device.
The best way would be that system
manufacturers develop a standard for fine-grained
privacy settings and their privacy friendly
distribution. But they seem not motivated enough to
do so. So states (governments) or confederation of
states should help their citizens and enforce such a
concept by law. It remains to be seen, whether the
concept of privacy by design and by default
combined with the increase of sanctions by the
upcoming GDPR will have the effect to foster such
solutions. Until policies like this exists we want to
help users to describe and enforce their privacy
preferences. Based on the requirements described
above, we have developed a concept for a distributed
privacy management solution, named PRIVACY-
AVARE. In the following, we first refine the basic
functional requirements and then describe a system
concept that implements these refined
functionalities. We also describe in more detail, how
such a system can account for the defined
compliance and usability requirements.
3.1 Refined Functional Requirements
In order to enhance privacy our software application
PRIVACY-AVARE will have the following three
essential functionalities:
(1) Enter the user’s preference profile:
PRIVACY-AVARE can be used to record the user’s
privacy preferences. The user of PRIVACY-
AVARE is supported by suitable explanations for
technical and legal laypersons. It creates a personal
preference profile. Therefore, we need a user-
friendly comprehensive GUI (see section 3.5) and
local data storage capabilities.
(2) Distribute the user’s preference profile: The
preference profile can be distributed via a central
service to all devices of the user. In order to secure
the exchange, a technical requirement is an end-to-
end encryption. Therefore, the user uses a locally
created symmetric key to encrypt the preferences
Figure 1: Different mechanisms for right management (Alpers et al., 2017b).
Citizen Empowerment by a Technical Approach for Privacy Enforcement
591
(e.g. using AES-256). The key is distributed to other
devices by embedding it into a QR code, displayed
on the first device and photographed by the second
one. This means that the user does not have to
entrust his preferences to a central service in plain
text; the key itself is not known to the central
service.
(3) Enable the user to control data access:
PRIVACY-AVARE enables the user to allow fine-
grained data access. Therefore, PRIVACY-AVARE
has different levels of data access control. Data
access can be blocked or filtered. Furthermore,
PRIVACY-AVARE provides the possibility to use
substitute data (no data or specially generated data)
in case the app stops working otherwise. This leads
to the following technical requirements:
a) PRIVACY-AVARE has to monitor data
access requests at runtime and block data
flows corresponding to the blockage rules
set by the user (privacy profile).
b) PRIVACY-AVARE has to extract data from
resources (e.g. address book) as defined by
filtering rules at runtime.
c) PRIVACY-AVARE has to deliver substitute
data to an application that would react with
failure to a denied data access otherwise.
Thus, PRIVACY-AVARE has to be able to
generate plausible substitute data.
d) PRIVACY-AVARE has to incorporate
permission settings of existing applications.
Those data flow filtering and blockage mechanisms
are executed during runtime.
3.2 System Concept
From the refined technical requirements, we derive
the following architecture: Figure 2 shows an
overview of the operating principle.
PRIVACY-AVARE is based on client server
architecture. The server is responsible for storage
and delivery of encrypted privacy profiles. The
client enables the user to set his privacy preferences
in three different levels of granularity. The privacy
settings result in rules for data flow control.
Furthermore, the client enforces these privacy data
flow rules. It controls which data flows from
hardware, sensors or other data sources to an app
using blocking, filtering or substitute data. The
client’s architecture is designed independently from
specific platforms (i.e. Android, iOS, Windows).
This facilitates the usage of PRIVACY-AVARE on
several devices (mobiles, tablets, smart homes, cars,
Figure 2: Operating Principle of PRIVACY-AVARE (Alpers et al., 2017b).
CLOSER 2018 - 8th International Conference on Cloud Computing and Services Science
592
etc.) with different operating systems in various
versions.
3.3 Compliance
As it should not be the obligation of the user to
decide on complex legal questions, our concept
addresses potential legal infringements. First, the
chosen technical solution should not interfere with
copyrights. Computer programmes are mostly
protected as long as the programming required an
act of certain creativity. While the functionality of a
programme is not protected, there is a dispute
whether only the programme code itself or also the
programme routine is protected under copyright law,
so that every change of programme sequences would
require a licence by the author or an exemption by
law (Spindler, 2012). Regarding Android, alterations
are permitted under the Apache licence, except
manufacturers’ modifications. The privacy solution
might alter such programme components, which fall
within an open source licence, so that other
proprietary components are just not triggered. Thus,
this could be considered as an interference with the
function of a programme, but not an alteration of the
proprietary code (KG Berlin 17.03.2010 - 24 U
117/08, 2010). Even if an infringement of code is
inevitable, an exemption by law provides the
possibility to alter computer programmes in order to
achieve the intended usability, which is especially
necessary, if flaws in the programming must be
corrected. A “correction” of data protection
infringements could be subsumed under this
provision (Bodden et al., 2013), but still legal
uncertainty remains.
If modifications of the operating system require
users to root their device, they face the risk of losing
the guarantee or warranty. While warranty claims
have to be fulfilled under the respective legal
obligations unless the defect was the result of the
modification, a voluntarily provided guarantee can
be revoked in case users violate contractual
conditions (as long as these conditions comprise no
unreasonable disadvantage to the customer). So,
from a user’s point of view rooting comprises
certain drawbacks which should be avoided.
The best solution to prevent copyright
infringements would be to avoid permanent
modifications of the proprietary code and to limit
alterations of the programme routine to a minimum,
using existing interfaces redirecting communication
between app and operating system.
Regarding substitute data, potential legal
infringements should be minimized by using
escalation steps. Even if apps are provided "for
free", users enter a contractual relationship. Whether
or not users risk to violate rights of their contractual
partners when they provide wrong information,
depends on the individual case. A “right to lie”
might apply in case of data protection infringements,
but a privacy solution should not require the user to
evaluate the legality of data collection as this would
require certain legal skills. Thus, the solution should
minimize implications by design choices. Only if the
blocking of data access leads to a loss in the
functionality of the app, empty data is provided (like
an empty address book and calendar, no sound, etc.),
so that the opponent cannot learn anything (wrong)
from such data. If the app detects this protective
measure, substitute data consisting of publicly
available information is provided, in order to reveal
no personal information about the user or third users
and meanwhile reduce potential damages due to
false data. Substitute data could be e.g. public
holidays (calendar), public authorities / companies
(address book), background noise (microphone),
image noise (camera). Special cases are location
data and IDs, as providing false location data could
also lead to negative consequences for other users,
e.g. when data is used by the app provider for traffic
jam prediction. To solve this, we propose a solution
that does not provide false location data, but blurred
substitute locations. These substitute locations are
randomly selected within a given radius. This radius
can then be given to the app provider as a
corresponding uncertainty. The app provider will
only receive the substitute location and uncertainty
information information that is not false, but
blurred.
3.4 Categorization
As Kelley et al. (2012) have shown, smartphone
users have difficulties to understand the implications
of their privacy settings. The goal of PRIVACY-
AVARE is therefore to reduce the complexity of
privacy decisions. We propose a categorization of
apps with similar profile of privacy settings based on
expert judgements. For example, one category
consists of applications providing navigation
functionality. Apps within the category navigation
need the location of the user, but can generally be
restricted, say, when accessing other sensor data.
Apps can be sorted within these predefined
categories, reducing the cognitive load for the user.
This way, the user does not need to make a decision
for every single app and every single privacy setting.
The predefined categorization can also be supported
Citizen Empowerment by a Technical Approach for Privacy Enforcement
593
by explanatory text to teach the user about sensible
settings regarding possible privacy risks. To provide
the necessary flexibility for apps that do not fit
within one category, the user still has the possibility
to overwrite specific settings for one app or to create
custom categories.
3.5 UI Prototype
We have developed an early UI prototype to
illustrate our idea. Initially, the user is shown a list
of predefined app-categories, as explained above.
Figure 3 shows exemplary screenshots of the
available settings within one category. On the left
side, the user has selected the app-category “Social
Networks and Communication” (containing Apps
like, for example, Facebook or WhatsApp). The
currently selected tab lists all available privacy
settings, like access to personal files, location,
sensors, contacts, and so on. Another tab “Apps”
allows adding and removing apps from the category,
which is not shown here.
The screenshot on the right in Figure 3 shows an
example, how the user could manipulate the settings
for contacts within one app-category (similar
possibilities are available for the other privacy
settings). The user may choose to block access to
contacts altogether, or to provide substitute date, in
case the app does not work without access to
contacts. As explained in section 3.3, this could be
addresses of public authorities or companies. The
user may also choose to filter data access. In this
example, the user decides to share the contact data
of “Gunther Schiefer” and “Stefanie Betz”, but to
hide the contact data of “Max Mustermann” and
“Martina Musterfrau”. These settings are then
applied to all apps within the category “Social
Networks and Communication”. Additionally, there
is the possibility for the user to create custom
categories or to override the settings for one app
within a category (not shown in the screenshots).
Figure 3: Screenshots of UI Prototype.
CLOSER 2018 - 8th International Conference on Cloud Computing and Services Science
594
4 SUMMARY AND OUTLOOK
In this position paper, we presented an approach
(PRIVACY-AVARE) to empower the privacy
enforcements of citizens. The approach is based on
an analysis of existing solutions for privacy
enhancement and the main functional requirements
are presented. We presented the general operating
principle of PRIVACY-AVARE and discussed some
non-functional compliance and usability
requirements. PRIVACY-AVARE enables users to
gain control over their data and thus enhances
confidence in cloud based services.
Currently, we are implementing our approach
(for Android devices) with a focus on privacy
enforcement for German citizens. The code is Open
Source (Apache 2.0) and available on GitHub
1
.
ACKNOWLEDGEMENTS
This work has been financed by the Baden-
Württemberg Stiftung gGmbH within the project
‘AVARE’.
REFERENCES
Alpers, S., Betz, S., Fritsch, A., Oberweis, A., Pieper, M.,
Schiefer, G., Wagner, M., 2017a. Existing Software
[WWW Document]. URL http://projects.aifb.kit.edu/
avare/existing_software/ (accessed 24.01.2018).
Alpers, S., Oberweis, A., Pieper, M., Betz, S., Fritsch, A.,
Schiefer, G., Wagner, M., 2017b. PRIVACY-
AVARE: An Approach to Manage and Distribute
Privacy Settings, in: 3rd IEEE International
Conference on Computer and Communications. pp.
14601468.
Backes, M., Bugiel, S., Hammer, C., Schranz, O., Von
Styp-Rekowsky, P., 2015. Boxify: Full-fledged App
Sandboxing for Stock Android, in: Proceedings of the
24th USENIX Conference on Security Symposium,
SEC’15. USENIX Association, Berkeley, CA, USA,
pp. 691706.
Bishop, M., 2012. Computer security: art and science.
Addison-Wesley, Boston ; Munich
Bodden, E., Rasthofer, S., Richter, P., Roßnagel, A., 2013.
Schutzmaßnahmen gegen datenschutz- unfreundliche
Smartphone-Apps: Technische Möglichkeiten und
rechtliche Zulässigkeit des Selbstdatenschutzes bei
Apps. Datenschutz Datensicherheit - DuD 37, 720
725. https://doi.org/10.1007/s11623-013-0300-1
1
https://github.com/fzi-forschungszentrum-
informatik/PRIVACY-AVARE
Christl, W., Spiekermann, S., 2016. Networks of control: a
report on corporate surveillance, digital tracking, big
data & privacy. Facultas, Wien.
Felt, A.P., Chin, E., Hanna, S., Song, D., Wagner, D.,
2011a. Android permissions demystified, in:
Proceedings of the 18th ACM Conference on
Computer and Communications Security. ACM, pp.
627638.
Felt, A.P., Greenwood, K., Wagner, D., 2011b. The
effectiveness of application permissions, in:
Proceedings of the 2nd USENIX Conference on Web
Application Development. pp. 77.
Forum Privatheit, 2014. White Paper Selbstdatenschutz.
Goldberg, I., Wagner, D., Thomas, R., Brewer, E.A.,
1996. A Secure Environment for Untrusted Helper
Applications Confining the Wily Hacker, in:
Proceedings of the 6th Conference on USENIX
Security Symposium, Focusing on Applications of
Cryptography - Volume 6, SSYM’96. USENIX
Association, Berkeley, CA, USA, pp. 11.
Kelley, P.G., Consolvo, S., Cranor, L.F., Jung, J., Sadeh,
N., Wetherall, D., 2012. A conundrum of permissions:
installing applications on an android smartphone, in:
International Conference on Financial Cryptography
and Data Security. Springer, pp. 6879.
KG Berlin 17.03.2010 - 24 U 117/08, 2010.
Leibenger, D., Möllers, F., Petrlic, A., Petrlic, R., Sorge,
C., 2016. Privacy Challenges in the Quantified Self
Movement An EU Perspective. Proc. Priv.
Enhancing Technol. 2016.
https://doi.org/10.1515/popets-2016-0042
Local Court Bad Hersfeld 15.05.2017 - F 120/17 EASO,
2017.
Schreiner, M., Hess, T., 2015. Examining the role of
privacy in virtual migration: The case of whatsapp and
threema.
Smith, E., 2010. iPhone applications & privacy issues: An
analysis of application transmission of iPhone unique
device identifiers (UDIDs). URL Www Pskl Uswpwp-
Contentuploads201009iPhone-Appl.-Priv.-Issues Pdf.
Spindler, G., 2012. Grenzen des Softwareschutzes.
Comput. Recht 28. https://doi.org/10.9785/ovs-cr-
2012-417
Vervier, L., Zeissig, E.-M., Lidynia, C., Ziefle, M., 2017.
Perceptions of Digital Footprints and the Value of
Privacy: SCITEPRESS - Science and Technology
Publications, pp. 8091.
https://doi.org/10.5220/0006301000800091
Citizen Empowerment by a Technical Approach for Privacy Enforcement
595