Abstract

A record of one’s relationships holds information about the person who keeps it and about everyone they know. This article reviews four bodies of evidence. First, privacy judgments depend on the context, the type of data and its use, and general privacy attitudes predict them poorly (Acquisti et al., 2015; Martin & Shilton, 2016). Second, people rarely attend to permission requests or read privacy policies, while information made salient at the moment of decision changes what they choose (Felt et al., 2012; Obar & Oeldorf-Hirsch, 2020; Tsai et al., 2011). Third, trust in the party that asks shapes whether people disclose, and can outweigh perceived privacy risk (Joinson et al., 2010; Dinev & Hart, 2006). Fourth, a smartphone’s data is exposed to friends and others nearby, and people’s locking choices track their risk perceptions (Muslukhov et al., 2013; Egelman et al., 2014). On this evidence, Mars combines a privacy lock with a 6-digit PIN, Face ID or Touch ID on iOS, or a browser passkey; imports only the contacts you pick; asks for iOS permissions one at a time; connects Gmail read-only; and reads public pages as a signed-out visitor. We also state the limits: the privacy paradox is contested, more control can raise disclosure (Kokolakis, 2017; Brandimarte et al., 2013), and this evidence supports the design direction without being a test of Mars itself.

Keywords: information privacy · contextual integrity · privacy paradox · control paradox · app permissions · trust and self-disclosure · smartphone locking · privacy lock · personal CRM

1. Introduction

A relationship assistant is useful only if people tell it things: who they met, what was said, what a friend is going through. That is information about the user and about third parties. A review in Science organised the empirical research on privacy behaviour around three themes: people are uncertain about the consequences of privacy-related behaviour and about their own preferences, their concern depends on context, and that concern is malleable, open to manipulation by commercial and governmental interests (Acquisti et al., 2015). Research on e-commerce argues that trust helps consumers overcome perceptions of risk and insecurity and makes them comfortable sharing personal information (McKnight et al., 2002).

This article asks under what conditions people are willing to record their relationships in a digital tool, and how those conditions translate into the design of a relationship assistant.

Method: this is a narrative literature review. We favoured peer-reviewed work accessible through a DOI, preferred meta-analyses, systematic reviews, experiments and field studies, and included work that limits or contests popular claims about privacy behaviour. Every factual statement carries its source, and the full reference list with original links follows the article.

2. Privacy depends on context, not on secrecy alone

Contextual integrity asks whether an information flow meets the privacy expectations of its context, and those expectations vary with the type of data collected and the context of use (Martin & Shilton, 2016). Martin and Shilton measured this with a factorial vignette survey that varied real-world contexts (such as medical, navigation and music), data types and data uses on mobile devices. Individuals’ general privacy preferences were of limited significance for predicting their judgments in specific scenarios, and common practices such as harvesting and reusing location data, images and contact lists did not meet users’ privacy expectations (Martin & Shilton, 2016).

Concern itself has structure. In two field surveys with 742 household respondents, Malhotra and colleagues found that Internet users’ information privacy concerns form three dimensions, collection, control and awareness, and that a model centred on these concerns explained a large amount of variance in the intention to release personal information when asked (Malhotra et al., 2004). Because concern depends on context (Acquisti et al., 2015), what matters is less whether a tool holds data than which data it collects, from where and for what use.

Following this evidence, Mars keeps each flow narrow. In Connected Platforms you tick the Google or iPhone contacts you want instead of importing your whole address book. When you send a link or a search finds a page, Mars reads what any signed-out visitor sees: no account, cookie, password or saved login goes with it, and nothing stays signed in (Martin & Shilton, 2016; Malhotra et al., 2004).

3. Permissions people can actually see and choose

Permission requests often fail at the moment they are shown. In an Internet survey of 308 Android users and a laboratory study with 25, only 17% of participants paid attention to permissions during installation, and only 3% of survey respondents answered all three comprehension questions correctly, although a notable minority showed both awareness and reasonable comprehension (Felt et al., 2012). Policies fare no better: when 543 participants joined a fictitious social networking service, 74% skipped the privacy policy through a “quick join” option, and 98% missed clauses about sharing data with the NSA and employers (Obar & Oeldorf-Hirsch, 2020).

Information presented where the decision is made does change choices. When a shopping search engine displayed privacy information clearly and compactly, consumers tended to buy from retailers who better protected their privacy, and some were willing to pay a premium to do so (Tsai et al., 2011). Among Android users deciding whether to download an app, perceived permission sensitivity raised privacy concerns, and a justification for the permission lowered them only for users with less experience of being a privacy victim (Gu et al., 2017).

Ongoing control also matters. In a field study, participants who had used an app permission manager for a week and then received nudges showing how often apps accessed their sensitive data 95% reassessed their permissions, and 58% restricted some of them further (Almuhimedi et al., 2015). Control has a known side effect, however: Brandimarte and colleagues tested, and their findings support, the hypothesis that more perceived control over the release of private information increases willingness to disclose sensitive information, which can leave people more vulnerable (Brandimarte et al., 2013).

Mars therefore asks for each kind of data separately. On iOS you grant Contacts, Calendar, Apple Health and Fitness one at a time, and Gmail is connected read-only under Settings > Connected Platforms, where you can disconnect it at any time (Felt et al., 2012; Gu et al., 2017; Almuhimedi et al., 2015).

4. Trust and clear limits on what a tool does

Joinson and colleagues measured privacy concern and, six weeks later, made an intrusive request for personal information (n = 759). Dispositional privacy concern, trust in the requester and perceived privacy of the request all predicted whether participants complied, and the effect of perceived privacy was mediated by trust. In a second experiment (n = 180), situational privacy and trust interacted: high trust compensated for low privacy, and vice versa (Joinson et al., 2010).

In a survey of 369 respondents, Dinev and Hart found that Internet privacy concerns inhibited e-commerce transactions, while Internet trust and personal Internet interest together could outweigh privacy risk perceptions in the decision to disclose (Dinev & Hart, 2006). McKnight and colleagues showed that trust is multidimensional, spanning disposition to trust, institution-based trust, trusting beliefs and trusting intentions, with 16 measurable subconstructs (McKnight et al., 2002).

Mars therefore states what it will not do. With Gmail connected read-only, Mars can search and read the mail you ask about in chat but never edits or sends a message; Mars drafts a reply and you copy, edit and send it yourself. Mars never asks for your passwords or cookies, never drives a browser for you, and when a page asks for a login or a verification check, it tells you it could not read it instead of guessing (Joinson et al., 2010; Dinev & Hart, 2006).

5. The people nearby

Threats to the data on a phone do not come only from distant attackers. In interviews with 22 people and a survey of 724, Muslukhov and colleagues found that users are generally concerned about insiders, such as friends, accessing their smartphone data; 12% reported a negative experience with unauthorized access, younger users were at higher risk, and the authors proposed an adversarial model that includes insiders for the design of data protection and authentication (Muslukhov et al., 2013).

Egelman and colleagues combined 28 interviews on why people do or do not use locking mechanisms such as PINs with two online experiments. Use of security features correlated strongly with risk perceptions, which indicates rational behaviour, yet most users likely underestimated the extent to which the data on their smartphones pervades their identities online and offline (Egelman et al., 2014).

Following this, Mars has a lock of its own. Set a 6-digit privacy PIN under Settings > Privacy Lock and Mars asks you to verify every time it opens, even on an unlocked or shared device; use Face ID or Touch ID on iOS, or register a passkey in your browser (Muslukhov et al., 2013; Egelman et al., 2014).

6. From evidence to design

These findings suggest that people are more willing to disclose to a tool when its information flows fit the context, when the choices are visible at the moment they are made, when the tool’s limits are clear enough to trust, and when the data stays out of sight of the people nearby. Mars meets each requirement as follows:

1. Expectations depend on data type and context → import only the contacts you pickIn Connected Platforms, tick the Google or iPhone contacts you want instead of importing your whole address book, so contact lists, a flow users judged not to meet their expectations, stay under your choice (Martin & Shilton, 2016; Malhotra et al., 2004).
2. Permissions are rarely read at install → one permission at a timeOn iOS, grant Contacts, Calendar, Apple Health and Fitness one at a time instead of in one sweeping request (Felt et al., 2012; Obar & Oeldorf-Hirsch, 2020; Gu et al., 2017).
3. Salient, specific limits shape choices → a read-only Gmail connectionGmail is connected read-only under Settings > Connected Platforms: Mars searches and reads the mail you ask about but never edits or sends a message, and you can disconnect it at any time (Tsai et al., 2011; Almuhimedi et al., 2015).
4. Trust compensates for perceived risk → nothing goes out in your nameMars drafts a reply and you copy, edit and send it yourself. It reads public pages as a signed-out visitor, takes no cookies or passwords, and tells you when it could not read a page instead of guessing (Joinson et al., 2010; Dinev & Hart, 2006; McKnight et al., 2002).
5. Insiders are a real risk → a privacy lock on every openWith a 6-digit privacy PIN under Settings > Privacy Lock, Mars asks you to verify every time it opens, even on an unlocked or shared device; use Face ID or Touch ID on iOS, or a passkey in your browser (Muslukhov et al., 2013; Egelman et al., 2014).

7. Limitations

First, the studies cited here test how people make privacy decisions, not Mars itself. They support our design direction but do not show that Mars is secure against any particular threat or that its users disclose more safely; those questions need their own evaluation, and this article makes no security claim beyond the product’s documented behaviour.

Second, the popular “privacy paradox”, the idea that people say they care about privacy but act otherwise, is contested. A review found studies that support the gap and studies that challenge it, and explained the diversity by differences in methods, contexts and definitions (Kokolakis, 2017). A meta-analysis of 166 studies from 34 countries (n = 75,269) found that privacy concerns did not predict social network site use, but that concerned users were less likely to use online services and share information and more likely to take protective measures (Baruh et al., 2017). A systematic review found strong support for the privacy calculus, with expected benefits among the best predictors of disclosure, and noted that many studies measure intentions rather than behaviour (Gerber et al., 2018).

Third, control is not a licence to disclose more. More perceived control can raise willingness to share sensitive information (Brandimarte et al., 2013), and trust can compensate for low privacy (Joinson et al., 2010). Mars’s choices and limits are meant to make data flows visible, not to encourage people to record more than they wish.

Fourth, several studies draw on specific samples or settings, such as Android users (Felt et al., 2012; Gu et al., 2017), a fictitious social networking service (Obar & Oeldorf-Hirsch, 2020) or self-reported intentions (Dinev & Hart, 2006; Malhotra et al., 2004). Generalizing to other cultures, platforms and to relationship records calls for care.

8. Conclusion

The evidence suggests that people are more willing to record their relationships when the information flow fits the context, when the choice is visible as it is made, when the tool’s limits are clear enough to trust, and when the people nearby cannot simply look. Mars’s privacy lock, selective contact import, one-by-one iOS permissions, read-only Gmail connection and signed-out web reading are designed on this evidence.

References

Listed in APA 7th edition style, each with its original DOI link.

1. Acquisti, A., Brandimarte, L., & Loewenstein, G. (2015). Privacy and human behavior in the age of information. Science, 347(6221), 509–514. https://doi.org/10.1126/science.aaa1465
2. Almuhimedi, H., Schaub, F., Sadeh, N., Adjerid, I., Acquisti, A., Gluck, J., Cranor, L. F., & Agarwal, Y. (2015). Your location has been shared 5,398 times! A field study on mobile app privacy nudging. Proceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems (CHI ’15), 787–796. https://doi.org/10.1145/2702123.2702210
3. Baruh, L., Secinti, E., & Cemalcilar, Z. (2017). Online privacy concerns and privacy management: A meta-analytical review. Journal of Communication, 67(1), 26–53. https://doi.org/10.1111/jcom.12276
4. Brandimarte, L., Acquisti, A., & Loewenstein, G. (2013). Misplaced confidences: Privacy and the control paradox. Social Psychological and Personality Science, 4(3), 340–347. https://doi.org/10.1177/1948550612455931
5. Dinev, T., & Hart, P. (2006). An extended privacy calculus model for e-commerce transactions. Information Systems Research, 17(1), 61–80. https://doi.org/10.1287/isre.1060.0080
6. Egelman, S., Jain, S., Portnoff, R. S., Liao, K., Consolvo, S., & Wagner, D. (2014). Are you ready to lock?. Proceedings of the 2014 ACM SIGSAC Conference on Computer and Communications Security (CCS ’14), 750–761. https://doi.org/10.1145/2660267.2660273
7. Felt, A. P., Ha, E., Egelman, S., Haney, A., Chin, E., & Wagner, D. (2012). Android permissions: User attention, comprehension, and behavior. Proceedings of the Eighth Symposium on Usable Privacy and Security (SOUPS ’12), 1–14. https://doi.org/10.1145/2335356.2335360
8. Gerber, N., Gerber, P., & Volkamer, M. (2018). Explaining the privacy paradox: A systematic review of literature investigating privacy attitude and behavior. Computers & Security, 77, 226–261. https://doi.org/10.1016/j.cose.2018.04.002
9. Gu, J., Xu, Y. (C.), Xu, H., Zhang, C., & Ling, H. (2017). Privacy concerns for mobile app download: An elaboration likelihood model perspective. Decision Support Systems, 94, 19–28. https://doi.org/10.1016/j.dss.2016.10.002
10. Joinson, A. N., Reips, U.-D., Buchanan, T., & Paine Schofield, C. B. (2010). Privacy, trust, and self-disclosure online. Human–Computer Interaction, 25(1), 1–24. https://doi.org/10.1080/07370020903586662
11. Kokolakis, S. (2017). Privacy attitudes and privacy behaviour: A review of current research on the privacy paradox phenomenon. Computers & Security, 64, 122–134. https://doi.org/10.1016/j.cose.2015.07.002
12. Malhotra, N. K., Kim, S. S., & Agarwal, J. (2004). Internet users’ information privacy concerns (IUIPC): The construct, the scale, and a causal model. Information Systems Research, 15(4), 336–355. https://doi.org/10.1287/isre.1040.0032
13. Martin, K., & Shilton, K. (2016). Putting mobile application privacy in context: An empirical study of user privacy expectations for mobile devices. The Information Society, 32(3), 200–216. https://doi.org/10.1080/01972243.2016.1153012
14. McKnight, D. H., Choudhury, V., & Kacmar, C. (2002). Developing and validating trust measures for e-commerce: An integrative typology. Information Systems Research, 13(3), 334–359. https://doi.org/10.1287/isre.13.3.334.81
15. Muslukhov, I., Boshmaf, Y., Kuo, C., Lester, J., & Beznosov, K. (2013). Know your enemy: The risk of unauthorized access in smartphones by insiders. Proceedings of the 15th International Conference on Human-Computer Interaction with Mobile Devices and Services (MobileHCI ’13), 271–280. https://doi.org/10.1145/2493190.2493223
16. Obar, J. A., & Oeldorf-Hirsch, A. (2020). The biggest lie on the Internet: Ignoring the privacy policies and terms of service policies of social networking services. Information, Communication & Society, 23(1), 128–147. https://doi.org/10.1080/1369118X.2018.1486870
17. Tsai, J. Y., Egelman, S., Cranor, L., & Acquisti, A. (2011). The effect of online privacy information on purchasing behavior: An experimental study. Information Systems Research, 22(2), 254–268. https://doi.org/10.1287/isre.1090.0260

Cite this article

Mars Research. (2026). Why people dare to record their relationships: privacy, control and trust. MarsDataAI. https://marsdata.ai/research/privacy-control-and-trust