You are currently viewing a new version of our website. To view the old version click .
Electronics
  • Article
  • Open Access

13 January 2022

Understanding Users’ Behavior towards Applications Privacy Policies

,
,
and
1
Department of Computer Software Engineering, University of Engineering and Technology, Mardan 23200, Pakistan
2
Department of Computer Science and Engineering, Hanyang University, Seoul 04763, Korea
3
Electrical and Space Engineering, Department of Computer Science, Luleå Tekniska Universitet, 97187 Lulea, Sweden
*
Author to whom correspondence should be addressed.
This article belongs to the Section Computer Science & Engineering

Abstract

Recently, smartphone usage has increased tremendously, and smartphones are being used as a requirement of daily life, equally by all age groups. Smartphone operating systems such as Android and iOS have made it possible for anyone with development skills to create apps for smartphones. This has enabled smartphone users to download and install applications from stores such as Google Play, App Store, and several other third-party sites. During installation, these applications request resource access permissions from users. The resources include hardware and software like contact, memory, location, managing phone calls, device state, messages, camera, etc. As per Google’s permission policy, it is the responsibility of the user to allow or deny any permissions requested by an app. This leads to serious privacy violation issues when an app gets illegal permission granted by a user (e.g., an app might request for granted map permission and there is no need for map permission in the app, and someone can thereby access your location by this app). This study investigates the behavior of the user when it comes to safeguarding their privacy while installing apps from Google Play. In this research, first, seven different applications with irrelevant permission requests were developed and uploaded to two different Play Store accounts. The apps were live for more than 12 months and data were collected through Play Store analytics as well as the apps’ policy page. The preliminary data analysis shows that only 20% of users showed concern regarding their privacy and security either through interaction with the development team through email exchange or through commenting on the platform and other means accordingly.

1. Introduction

As we all know, smartphones have become the most essential part of our life in recent years: everyone is familiar with the usage of smartphones, and there are a lot of smartphones in the market of different operating systems, such as Android and iOS. However, there are more Android than iOS users. This is because Android is very flexible and open source, and a lot of new applications are available and developed each day in Android. People use Android smartphones for different purposes, including for communication, social media, entertainment, gaming, camera, and many other kinds of activities. Most Android users use and install third-party applications from Android Marketplace (Google Play), Amazon Appstore, and many other kinds of third-party app stores. As the Android mobile phones become widespread and powerful, many Android applications are collecting more and more sensitive data from their users through sensors, and this can be done maliciously [1].
The majority of the Android applications require permissions from users to receive access to smartphone resources when interacting with it. These permissions contain can be classified as follows. Normal permissions include battery status and internet. Alarm, time zone, vibrate, wake lock, and many more of these are permissions with low or minimum risk. Dangerous permissions include camera permission, messages read permission, record permission, storage permission, location permission accounts permission, phone gallery, etc., which fall in the dangerous category because they are riskier and can cause user data leakage if allowed. “Signature” means the Android application only gives this permission for asking permission, which is signed with a certificate. “Signature or System” means the system allows permission at the time of installation. However, these work only when the user attempts to use the application.
In this regard, the Google Play store made it compulsory for a developer to add a privacy policy in such apps that take permission from their users. The privacy policy of the app describes their usage of the apps and how the app can collect the data from the user and the data’s uses. Such apps have permissions that restrict them from any malicious activities. While the application is installed, Google Play does not verify if the application is safe to use or not. Google Play relies on the bouncer, a dynamic environment to prevent itself from dangerous attacks. It cannot analyze the vulnerability of existing applications. The Google Play store thoroughly verify that the apps are safe or malicious, and according to the Google policies, it is the responsibility of the user to allow certain permissions to such applications or not. Most applications contain irrelevant permissions to the main features of apps and collect data from users that cause leaks of users’ private information and can harm them, and it has been observed that most people pay no attention to permissions while installing the applications and cause very serious problems for the users.
Apps that access user data and sensitive permissions must include a privacy policy within the app and a link in the app store listing page. This protects against threats. However, the main problem is that the user does not bother to read the privacy policy of the app and understand the purpose of such sensitive permissions. The rest of the paper is organized as follows, Section 2 briefly discusses related works in the literature, Section 3 explains the proposed researched methodology, Section 4 introduces the reflective process, Section 5 introduces outcome of undertaking coursework, Section 6 provides a brief discussion and recommendations, and finally Section 7 presents conclusions and future directions. This study will help Android users follow the Google’s privacy policy to read/visit the apps blog, and it will also help users to verify app permissions before installing apps to protect themselves from personal identity theft, banking and financial theft, credit card scams, and more.

3. The Research Method

Before describing the method, we must present the aim of the privacy policy in Android Apps. Apps that use such kinds of sensitive permission gain control of a device, steal private information from users, consume excessive battery, use telephone services to steal money from users’ bank accounts, and even to turn the device into a botnet zombie [18].
There are a variety of security issues on Android phones, such as unauthorized access from one app to the others (information leakage), permission escalation, repackaging apps to inject malicious code, colluding, and Denial of Service (DoS) attacks. Android applications and other applications are not allowed to give access to the resource architecture of Android. Before installing an application, the user must give access permission to the app. When the app receives the resource of an operating system, it results in the leakage of information. In this research, it has been proven that 80 to 90 percent of users did not read the privacy of the application [19,20].
Developers upload third-party applications on a daily basis to Google Play, but Google does not have the proper mechanism to ensure that users read the privacy policy before installing the applications.
Therefore, for this research, a total of seven applications were developed for the proposed study and uploaded to two different accounts. Additionally, irrelevant permissions to the main features of apps were added in these apps. To find the result of our target area of privacy-first, we created a private URL that was used to create a private blog; this private URL is compulsory in the app store listing for apps that have permissions. This blog describes the uses of permission in apps. A flag counter option is created in the privacy blog. When a user visits the page, its entry is counted with its location (country name). Secondly, a button is created inside our apps for the privacy policy on the main screen, which is hyperlinked to our URL privacy page. When pressing this button, the user is redirected to our privacy policy blog, and it will count the visitor. Major steps involved in the proposed study are shown in Figure 2.
Figure 2. Major steps involved in proposed study.
The conceptual representation of our study is shown in Figure 3. The data gathered in this exercise were fed into an Excel sheet to perform statistical analysis on the results about app downloads. In addition, apps’ download country, their population, and their literacy rates were also analyzed.
Figure 3. Conceptual representation of our study.

4. The Reflective Process

The ultimate objective of this study is to find the percentage of Android users who would allow irrelevant apps with permissions to sensitive resources on their devices and to find the percentage of Android users who would take the time to read/visit the privacy policy blog of the apps being installed. Another objective is to obtain a general overview of global Android users’ attitudes towards their devices and information privacy and correlate the collected data with the overall literacy rate country-wise. This ultimate objective may be achieved by answering the following questions.
  • Will Android users provide sensitive permissions to apps and install them on their devices?
  • Will Android users follow Google’s privacy policy to read/visit the app’s blog?
  • Will Android users understand the privacy risk posed by the apps mentioned in the blog contents and be willing to withdraw from the installation of the app?
Can Google/Android App Store automatically detect the security risks posed by the apps and remove/block the apps?
Once the outcomes of the above issues have been obtained, some suggestions are provided to enhance the app’s checking process at the time of uploading the app to the Google Play store.

5. Outcomes of Undertaking Coursework

As discussed in Section 2, a total of seven applications were developed for the proposed study; the following steps were followed for data collection and processing.
Seven applications were developed in Android Studio, and irrelevant permissions were added to each app in their manifest file that were not related to the main features of the apps. Figure 4 show one of the app’s permissions. These apps were uploaded to the Google Play store, and a valid privacy policy URL was provided in the app store listing. This privacy policy URL could be visited in two ways: one was from the app’s main page, in which a privacy policy button is provided to users to visit the app’s privacy policy URL; the other is that users could visit the privacy policy URL from the Google Play store. The privacy policy button on the app home page was connected to our server to find the total number of privacy policy visitors coming from apps after installing the application. Below, we can see the results and the findings of the data from the last two years.
Figure 4. Total number of apps downloads.
As mentioned, seven very popular apps were developed from two different accounts for the study. Figure 5 and Figure 6 show the total number of apps downloads from account 1, and Figure 7 and Figure 8 show app downloads from account 2.
Figure 5. App downloads from account 1.
Figure 6. App downloads from account 1.
Figure 7. Account 2’s app downloads.
Figure 8. Account 2’s app downloads.
Figure 9 shows the total number of users who visited the app by clicking the privacy button in the app. This picture was taken from the flag counter implemented in the blog as the % of blog visitors from each country.
Figure 9. Account 1’s total countries and privacy policy visitors.

6. Discussions and Recommendations

The seven chosen apps received a total of 25,041 app downloads and a total of 5739 privacy policy visitors. These numbers show that a much smaller number of people visit the privacy policy. Additionally, it is alarming to know that only 22% of users take privacy as a serious concern and the rest of them pay no attention, as shown in Figure 10. The privacy directly depends on the country’s literacy rate, with higher literacy rates being associated with more visitors. For example, the United States’ literacy rate is 86%, the total app downloads are 1232, and there were 600+ private visitors, which indicates that more than 50% of people visit the privacy policy. Similarly, if we take an example of Pakistan and India, which have lower literacy rates compared to the United States, the privacy policy visitors comprised 18% and 11%, respectively.
Figure 10. Overall privacy policy visitors. To test whether education had any relationship with how people handled their privacy, specifically the privacy of the content/data on their smartphones, the country-wise literacy rates data were obtained from [16]. These data were then augmented with information, such as country-wise Android app downloads and country-wise count of Android apps’ privacy page access data from our experiments. The list was then sorted based on the number of privacy page accesses country-wise to choose the sample data. Sample data are shown in Table 1, and the full data are shown in Appendix A.
Pearson correlation between the Policy Page Accesses variable and the literacy rate of countries was calculated to find any relationship between the general education of the people of a country and their behavior toward the privacy of their Android smartphones. The Pearson correlation coefficient r = −0.58 shows a relatively strong negative correlation among the two variables, indicating that the higher the literacy rate of a country, the less its people pay attention to the privacy of the data/content on their smartphones. The outcome seems a bit strange, but a possible explanation could be that the educated people put more trust in their service providers than the less educated. It also suggests that people rely on the Google Play store’s service to take care of the privacy of their data, even though Google expects them to be vigilant towards the permissions requests while installing Android applications.
Base on the presented results following two suggestions are being made:
(1)
Google needs to verify applications that manually remove spam apps at the time of uploading, as the iTunes store does.
(2)
A description should be added to run time permission that indicates the use of resources and needs verification.

7. Conclusions and Future Directions

In this paper, two different methods were used. First, we uploaded seven different apps on the Google Play store and analyzed them for 1 year. Second, we created a blog for a privacy policy. While developing the app, a “Privacy” button was created, which was a link to the blog. For these applications, we observed the number of users who visit the blog by clicking the privacy button and found that a very small number of users read the privacy blog. After obtaining this result, we concluded that there is no proper guidance about the privacy policy to the user from the Google Play store. The number of app downloads is greater more than number of privacy policy visitors as shown in Table 1. A privacy policy is a visitor also depends upon the language in which privacy is written more languages in the future can be added. Furthermore, as a future direction, we are planning to create a way in which the user must read the privacy policy of an application while installing and interacting with the application accordingly.
Table 1. Top countries with apps downloads with literacy rate and privacy policy visitors.

Author Contributions

Conceptualization, S.U.; Investigation, S.U., M.S.K. and M.H.; Supervision, M.S.K. and C.L.; Writing—original draft, S.U. and M.S.K. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the Korea Meteorological Administration Research and Development Program under Grant KMI 2021-01310.

Data Availability Statement

Data are available within the manuscript.

Conflicts of Interest

The authors declare no conflict of interest.

Appendix A

The following table shows the information about data collected and results. The country-wise literacy rate data were obtained from [16], and population data were obtained from [17].
Country NamePolicy Page AccessesTotal Download of Apps from That CountryLiteracy Rate of the Country %Total Population
United State607123286%323,995,528
Pakistan277148255%201,995,540
India442383769%1,266,883,598
Brazil12865490%5,823,665
South Africa368993%54,300,704
Turkey150131393%80,274,604
Philippines2722495%102,624,209
Lebanon2813690%6,237,738
Russia3526395%142,355,415
Mexico140107993%123,166,749
United Kingdom3526499%64,430,428
France3318699%66,836,154
Ireland154099%4,952,473
Egypt8046772%94,666,993
Japan329399%126,702,133
Iraq114100179%38,146,025
Algeria3250373%40,263,711
Kazakhstan525100%18,360,353
South Korea37102100%50,924,172
United Arab Emirates1610190%5,927,482
Australia57199%22,992,654
Dominican Republic325590%10,606,865
Malaysia2517293%30,949,962
Romania169799%21,599,736
Poland8108100%38,523,261
Canada1213499%35,362,905
Singapore53696%5,781,728
Netherlands107399%17,016,967
Ukraine25101100%44,209,733
Argentina3534798%43,886,748
Maldives53098%392,960
Bangladesh2537260%156,186,882
Jamaica3114388%2,970,340
Saudi Arabia5523994%28,160,273
Honduras146085%8,893,259
China2795%1,373,541,278
Israel14998%8,174,527
Indonesia16556593%258,316,051
Taiwan276898%23,464,787
Ghana11471%26,908,262
Austria81898%8,711,770
Finland819100%5,498,211
New Zealand3999%4,474,549
Zimbabwe274784%14,546,961
Germany319599%80,722,792
Greece278297%0,773,253
Trinidad and Tobago81999%1,220,479
Italy259099%62,007,540
Colombia166493%47,220,856
Azerbaijan30100100%9,872,765
Czech208699%10,644,842
Malawi14861%18,570,321
Rwanda11066%12,988,423
Norway45100%5,265,158
Tunisia2218979%11,134,588
Hungary23399%9,874,784
Uruguay41998%3,351,016
Spain2810198%48,563,476
Haiti42189049%10,485,800
Thailand4213196%68,200,824
Algeria2756473%40,263,711
Vietnam2710493%95,261,021
Tunisia2150379%11,134,588
Libya188490%6,541,948
Jordan1610293%8,185,384
Guinea154795%12,093,349
Cambodia1218474%15,957,223
Oman129887%3,355,262
Sri Lanka119591%22,235,000
Laos113484%7,019,073
Syria107886%17,185,170
Panama102494%3,705,246
Georgia972100%4,928,052
Somalia98539%10,817,354
Palestinian Territory82998%1,753,327
Nigeria921051%186,053,386
Venezuela75795%30,912,302
Bulgaria73398%7,144,653
Burma74989%56,890,418
Croatia72899%4,313,707
Reunion61099%66,836,154
Belgium62999%11,409,077
Slovakia617100%5,445,802
Yemen68068%27,392,779
El Salvador64284%6,156,670
Moldova51999%3,510,485
Macedonia53298%2,100,025
Guatemala56078%15,189,958
French Guiana53883%66,836,154
Afghanistan57032%33,332,025
Mozambique510151%25,930,150
Suriname5895%585,824
Peru42294%30,741,062
Hong Kong41499%7,167,403
Bosnia and Herzegovina43498%3,861,912
Belize44183%353,858
Ecuador46992%16,080,778
Puerto Rico42293%3,578,056
Switzerland31099%8,179,294
Uganda35173%38,319,241
Senegal34052%14,320,055
Kenya35972%46,790,758
Angola35571%20,172,332
Uzbekistan324100%29,473,614
Latvia312100%1,965,686
Bahrain31995%1,378,904
Serbia39598%7,143,921
Tanzania24768%52,482,726
Fiji2894%915,303
Guyana21185%735,909
Mongolia21198%3,031,330
Burkina Faso28229%19,512,533
Qatar22096%2,258,283
Slovenia212100%1,978,029
Brunei2496%436,620
Barbados24100%291,495
Armenia114100%3,051,250
Lithuania114100%2,854,235
Cyprus11199%1,205,575
French Polynesia1198%285,321
Togo13060%7,756,937
Paraguay11094%6,862,812
Kyrgyzstan1899%5,727,553
Belarus115100%9,570,376
Sweden11299%9,880,604
Sudan13174%36,729,501
Mauritania14146%3,677,293
Kuwait11796%2,832,776
Portugal12994%10,833,816
Costa Rica11797%4,872,543
Unknown—European Union192N/ANANA
Unknown—Asia/Pacific Region149N/ANANA
Unknown—Anonymous Proxy1N/ANANA
MontenegroN/A498%622,303
LaosN/A284%6,758,640
EstoniaN/A199%1,312,442
NamibiaN/A390%2,479,713
TongaN/A299%107,122
GuamN/A199%162,896
MadagascarN/A364%24,894,551
ChileN/A10996%17,909,754
DenmarkN/A199%5,707,000
BeninN/A538%10,872,298
AlbaniaN/A2297%2,926,348
AndorraN/A2100%77,281
ArubaN/A386%104,822
Total573925,041

References

  1. Gibler, C.; Crussell, J.; Erickson, J.; Chen, H. AndroidLeaks: Automatically detecting potential privacy leaks in android applications on a large scale. In International Conference on Trust and Trustworthy Computing; Springer: Berlin/Heidelberg, Germany, 2012; pp. 291–307. [Google Scholar]
  2. Koch, S.; Kerschbaum, M. Joining a Smartphone Ecosystem: Application Developers’ Motivations and Decision Criteria. Inf. Softw. Technol. 2014, 56, 1423–1435. [Google Scholar] [CrossRef]
  3. Number of Android Smartphone Users in the United States from 2014 to 2022. Available online: https://www.statista.com/statistics/330695/number-of-smartphone-users-worldwide/ (accessed on 11 August 2021).
  4. Xie, K.; Liang, B.; Dulebenets, M.A.; Mei, Y. The impact of risk perception on social distancing during the COVID-19 pandemic in China. Int. J. Environ. Res. Public Health 2020, 17, 6256. [Google Scholar] [CrossRef] [PubMed]
  5. Colizza, V.; Grill, E.; Mikolajczyk, R.; Cattuto, C.; Riley, S. Time to evaluate COVID-19 contact-tracing apps. Nat. Med. 2021, 27, 361–362. [Google Scholar] [CrossRef] [PubMed]
  6. Available online: http://www.android.com/market/ (accessed on 10 August 2021).
  7. Felt, A.P.; Chin, E.; Hanna, S.; Song, D.; Wagner, D. Android Permissions Demystified. In Proceedings of the ACM Conference on Computer and Communications Security, Hong Kong, China, 22–24 March 2011. [Google Scholar] [CrossRef] [Green Version]
  8. Reardon, J.; Feal, A.; Wijesekera, P. An Exploration of Apps’ Circumvention of the Android Permissions System. Available online: https://www.usenix.org/conference/usenixsecurity19/presentation/reardon (accessed on 10 August 2021).
  9. Android Developers Reference. Available online: http://developer.android.com/reference/ (accessed on 11 August 2021).
  10. Kovacs, E. Wi-Fi Direct Flaw Exposes Android Devices to DoS Attacks. Available online: https://www.securityweek.com/wi-fi-direct-flaw-exposes-android-devices-dos-attacks (accessed on 11 August 2021).
  11. Kelley, P.G.; Cranor, L.F.; Sadeh, N. Privacy as Part of the App Decision-Making Process. In Proceedings of the Conference on Human Factors in Computing Systems, Paris, France, 27 April–2 May 2013. [Google Scholar] [CrossRef] [Green Version]
  12. Kelley, P.G.; Consolvo, S.; Cranor, L.F.; Jung, J.; Sadeh, N.; Wetherall, D. A Conundrum of Permissions: Installing Applications on an Android Smartphone. In Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Springer: Berlin/Heidelberg, Germany, 2012; Volume 7398. [Google Scholar] [CrossRef] [Green Version]
  13. Balebako, R.; Marsh, A.; Lin, J.; Hong, J.; Faith Cranor, L. The Privacy and Security Behaviors of Smartphone App Developers; Carnegie Mellon University: Pittsburgh, PA, USA, 2014. [Google Scholar] [CrossRef]
  14. Rashidi, B.; Fung, C. A Survey of Android Security Threats and Defenses. J. Wirel. Mob. Netw. Ubiquitous Comput. Dependable Appl. 2015, 6, 3–5. [Google Scholar] [CrossRef]
  15. Zhou, W.; Zhou, Y.; Jiang, X.; Ning, P. Detecting Repackaged Smartphone Applications in Third-Party Android Marketplaces. In Proceedings of the 2nd ACM Conference on Data and Application Security and Privacy, San Antonio, TX, USA, 7–9 February 2012. [Google Scholar] [CrossRef]
  16. Sun, X.; Zhongyang, Y.; Xin, Z.; Mao, B.; Xie, L. Detecting Code Reuse in Android Applications Using Component-Based Control Flow Graph. In IFIP Advances in Information and Communication Technology; Springer: Berlin/Heidelberg, Germany, 2014; Volume 428. [Google Scholar] [CrossRef] [Green Version]
  17. Brinkhoff, T. City Population. Available online: https://www.citypopulation.de/en/world/bymap/LiteracyRates.html (accessed on 26 August 2021).
  18. Talal, M.; Zaidan, A.A.; Zaidan, B.B.; Albahri, O.S.; Alsalem, M.A.; Albahri, A.S.; Alamoodi, A.H.; Kiah, M.L.M.; Jumaah, F.M.; Alaa, M. Comprehensive Review and Analysis of Anti-Malware Apps for Smartphones. Telecommun. Syst. 2019, 72, 285–337. [Google Scholar] [CrossRef]
  19. Android. Google Android Documents, Android Application Sandboxing Mechanism. Available online: http://developer.android.com/training/articles/security-tips.html (accessed on 26 August 2021).
  20. Gunasekera, S. Android Apps Security; Springer: Berlin/Heidelberg, Germany, 2020. [Google Scholar] [CrossRef]
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Article Metrics

Citations

Article Access Statistics

Multiple requests from the same IP address are counted as one view.