Fig. 1: Electronic Frontier Foundation (eff.org) graphic created by EFF Senior Designer Hugh D'Andrade of a Laptop Spying on its User
Have you ever stopped yourself from posting a comment, typing in a search engine, or sharing a news article because you were worried about being tracked online? You are not alone. In fact, these actions are now so prevalent that researchers have begun linking it to a form of self-censorship called “the chilling effect”, i.e. people’s propensity to restrict their actions when they feel like they are being watched. If digital surveillance (AKA dataveillance) really is causing us to self-censor, we may be heading down a dangerous path; beyond being a threat to our freedom of expression, dataveillance could curtail the diversity of opinion that gives rise to new ideas. Given these potential risks, we must address these questions:
- Do we self-censor when watched?
- Why do we self-censor?
- Can empowering users reduce chilling effects?
- What methods are effective for reducing chilling effects?
Do We Self-Censor When Watched?
To answer the first question: yes, we censor ourselves when we are watched. Across multiple studies, researchers found that exposure to dataveillance made people less comfortable searching for information, sharing opinions, and disclosing personal details online. [3][4][6] Although the long-term implications of this are still unclear because there has yet to be a multi-year study, experts agree that being watched makes us disengage not only from active participation like speaking, but also passive activities like browsing.
Why Do We Self-Censor?
Although it may appear irrational for someone to refrain from legal activities out of a fear of being watched, there are actually good reasons for this. One of the leading motivations is the perceived risk of ambiguous laws. In April 2026, the UK made several major changes to its Online Safety Act: companies had to start reading private messages, and users had to scan their faces and IDs to verify their age. Predictably, this caused widespread chilling effects; when investigating the reasons behind this, researchers found that “the ambiguity surrounding the Online Safety Act (OSA) intensified these fears (punitive consequences), as individuals found it challenging to discern the boundaries of ‘legal but harmful’ speech.” [2] In other words, people self-censor in order to not accidentally break the law and provoke legal repercussions. Therefore, to ensure that people can continue to participate in online discourse, it is crucial that the law clearly distinguishes between legal and illegal.
Can Empowering Users Reduce Chilling Effects?
One proposed method of reducing chilling effects is to empower users. The thinking goes: by educating them on privacy literacy, they could proactively take control of their own data; as they feel more confident in their privacy, they would self-censor less. Although this sounds like a fine idea in theory, it may actually produce the opposite effect: higher privacy literacy could actually amplify chilling effects. This is largely due to a phenomenon called privacy cynicism. The more knowledgeable someone is about online privacy, the more likely they think protecting their privacy is futile [11]; they feel a weaker sense of privacy, and a greater sense of being watched; the more they perceive being monitored, the more they constrain their actions [15]. In other words, attempting to empower users by increasing their privacy literacy may not be the best way to prevent people from censoring themselves.
Fig. 2: Symbol for General Data Protection Regulation (GDPR) in Europe.
What Methods are Effective for Reducing Chilling Effects?
As previously mentioned, ambiguous laws are a root cause of chilling effects; consequently, some have called for legislation to address self-censorship. Professor Jonathan W. Penney, research fellow at the University of Toronto’s Munk School of Global Affairs and Public Policy, argues that regulators can reduce ambiguity by requiring governments and corporations to routinely disclose what data they have collected on you, and what that data can and cannot be used for in a legal context. [9][10]
Although the US legal system has not utilized any of Penney’s suggestions, the EU already implemented such protections with its General Data Protection Regulation (GDPR) and it appears to be effective. In a study comparing Dutch and American responses to hypothetical surveillance concerns, the former consistently reported lower chilling effects, demonstrating the efficacy of regulation in curbing self-censorship. Granted, legislators must be careful when drafting new acts, since more comprehensive privacy protection laws may have the unintended consequence of making citizens more complacent about guarding their data; that said, structural legal changes still appear to be much more effective than personal empowerment as a method of reducing chilling effects. Implementing these laws in addition to utilizing privacy tools like VPNs and encrypted messaging, which demonstrably lower chilling effects [12], could help with self-censorship.
Takeaways
The constant digital surveillance we are currently living under has undeniably negative effects on our freedom of expression. However, this does not mean that we should constantly censor ourselves or abandon the internet; you can contact your local representative, oppose anti-privacy legislation, call for greater privacy protections, download a VPN… The odds may be stacked in favor of surveillance, by demanding better governance and utilizing privacy tools, we can reclaim some of our privacy and information autonomy.
References
[1] Buchi, M., Festic, N., & Latzer, M. (2022). The chilling effects of digital dataveillance:
A theoretical model and an empirical research agenda. Big Data & Society. 8
[2] Daruwala, N. A. (2025). Social media, expression, and online engagement: A psycho-
logical analysis of digital communication and the chilling effect in the UK. Frontiers in Communication.
[3] Festic, N., Buchi, M., Latzer, M., & Odermatt, C. (2026). Dataveillance inhibits legiti-
mate communication: Causal evidence for chilling effects. Journal of Communication.
[4] Meier, Y., & Masur, P. K. (2026). Escaping the digital panopticon? Longitudinal effects
of dataveillance salience shocks on privacy attitudes and inhibited behaviors.
[5] Murray, D. (2024). Making tangible the long-term harm linked to the chilling effects of
AI-enabled surveillance: Can human flourishing inform human rights?
[6] Odermatt, C., Festic, N., Jaramillo-Dent, D., Kappeler, K., & Latzer, M. (2025). Trig-
gers of a sense of dataveillance: Empirical insights into characteristics and determinants.
[7] Online Safety Act 2023, 2023 c. 50, Office of Communications (2026).
[8] Penney, J. (2016). Chilling effects: Online surveillance and Wikipedia use. Minnesota
Law Review.
[9] Penney, J. W. (2022). Understanding chilling effects. Minnesota Law Review, 106(3),
1451-1513.
[10] Penney, J. W. (2025). The future of chilling effects and how to stop it. In Chilling
Effects: Repression, Conformity, and Power in the Digital Age (Chapter 10, pp. 168-
189). Cambridge University Press.
[11] Respi et al. (2026) Lower cynicism, not higher literacy, promotes protective behavior:
Exploring the ’privacy exception’ in the digital inequality framework.
[12] Sizov, A. (2026). ”Chilling effects” of state surveillance: Determinants of Russians’ po-
litical participation, self-censorship, and digital resistance. Higher School of Economics. 9
[13] Strycharz, J., & Segijn, C. M. (2023). Consumer differences in chilling effects from
dataveillance. In The Routledge Handbook of Digital Consumption. Routledge.
[14] Strycharz, J., & Segijn, C. M. (2024). Chilling Effects as a Result of Corporate Surveil-
lance in Digital Communication: A Comparison Between American and Dutch Media
Users. International Journal of Communication.
[15] Zhang, et al. (2025). Google knows me too well! Coping with perceived surveillance in
an algorithmic profiling context. Computers in Human Behavior. 10
