How Self-Censorship Is Unaliving Online Freedom

When was the last time you avoided using a particular word on social media, not because it was hateful or offensive, but because you were scared your post might get flagged, suppressed, or disappear from a platform altogether? 

If you’ve ever typed “pew pew” instead of “gun,” “unalived” instead of “killed,” “corn” instead of “pornography,” or “grape” instead of “rape,” you’ve already participated in one of the internet’s quietest but most influential trends: self-censorship.

This isn’t about people pushing boundaries or trying to say outrageous things online. It’s about ordinary users adapting their language to be heard at all. In my opinion, this should worry everyone and anyone who cares about internet freedom, especially in English-speaking countries such as the United States, Canada, the EU, and Australia.

We’re Editing Ourselves to Stay Visible

Self-censorship is typically defined as deliberately withholding or modifying your expressions out of fear of consequences rather than direct prohibition. It’s like driving way below the speed limit “just in case.” According to the Cambridge Dictionary, it means choosing not to say what you really think or want to say because of how others (including institutions) might react.

On today’s internet, the institutions mentioned above are oftentimes the social media platforms we’ve come to love and the algorithms that enforce their rules with an iron fist.

Research shows that users are now routinely altering their own language, spelling, and phrasing, whether in the videos they make or in the comment section, to avoid the wrath of automated moderating systems. On platforms like TikTok, the use of emojis (“🍇” for “rape”, “🔫” to discuss shootings), euphemisms (“unallived” for “killed”), or coded language (“🧻 people” for “white people”) is prevalent when discussing topics that may otherwise trigger content suppression algorithms, even if the discussion itself is warranted, good-hearted, and otherwise legitimate.

This kind of behavior isn’t accidental; it’s learned. Social media users have noticed that some of the words they used led to fewer views, lower user engagement, outright video removals, or repeated account warnings, and they adjusted their behavior accordingly, especially if they make a living by creating content on these platforms.

The now-ubiquitous terms, like “unalived”, “pew pew”, and “grape”, are some of the clearest examples of people straying from anything that may be considered inappropriate or harmful. This is especially common for the true-crime community, where creators review infamous crimes, their perpetrators, and their unfortunate victims. 

Yes, censoring yourself in favor of the algorithms is clever, but not only is it disrespectful to the victims and their families, but it’s also revealing the big picture issue that people are changing the way they speak, not because they want to; they’re doing it because they feel they have to, in order to appease the algorithm overlords. 

That, by definition, is self-censorship. Or “self 🤫”, if you’re so inclined.

Why It Happens: Algorithms, Ambiguity, and Fear

Social media moderation today is largely automated. Complex algorithms scan for keywords and patterns to enforce platform policies at scale. While these systems are designed to limit genuinely harmful content, they often struggle with context, nuance, and intent.

As a result, users are left guessing. When posts are downranked, demonetized, or removed without clear explanations, the lesson is simple: next time, avoid risk. Don’t say the word. Don’t touch the topic. Find a workaround.

A University of Illinois study found that people consciously modify their language to avoid having content suppressed, especially when discussing sensitive topics. Importantly, this behavior isn’t limited to extreme cases; it also affects everyday conversations about news events, personal experiences, and social issues.

Even regulatory efforts, such as the EU’s Digital Services Act, which aim to increase transparency in content moderation, reveal how inconsistent and opaque enforcement can be in practice. Much of what shapes online speech happens quietly, automatically, and without meaningful user recourse.

The result is an environment where people self-police, not because they understand the rules, but because they fear invisible penalties.

What This Means for You

On an individual level, self-censorship changes how people like you or me show up online. It discourages honesty and clarity. It makes people second-guess whether sharing a personal story, an opinion, or even a factual observation is worth the potential downside.

This effect isn’t evenly distributed. Marginalized communities are often more cautious about how they speak online because they’re more likely to experience moderation or backlash. When those voices self-censor more than others, public discourse becomes narrower and less representative.

Over time, people don’t just censor words; they avoid topics altogether. Conversations about mental health, violence, or controversial events become indirect, coded, or absent. That’s not because these discussions aren’t important, but because the cost of having them feels too high.

The Societal Cost of Self-Censorship

Self-censorship has long been recognized as a threat to free societies around the globe. Historically, it has kept journalists from reporting certain important stories and citizens from challenging power, not through force, but through fear of consequences.

When this dynamic plays out on a large scale on social media, the stakes are even higher than a single TikTok video. Platforms function as modern public squares. When people routinely soften, distort, or withhold their views to appease algorithms, the boundaries of acceptable speech become increasingly blurred without anyone formally drawing them.

What concerns me most is that this shift happens quietly. There’s no public debate about which words disappear or why specific conversations fade. Language simply adapts around unseen constraints, and over time, those constraints start to feel normal.

That’s how self-censorship becomes cultural. 

What This Means for Internet Freedom

Across the U.S., Canada, the EU, Australia, and other countries, freedom of expression is widely regarded as a core democratic value. That doesn’t mean speech should be consequence-free, nor does it mean platforms shouldn’t moderate genuinely harmful content.

But there’s a meaningful difference between moderation and an environment where users feel compelled to disguise their thoughts just to remain visible. When people speak in euphemisms to avoid punishment, the burden of censorship shifts from platforms to people like you and me. 

Instead of clear rules and transparent enforcement, we get uncertainty, and uncertainty breeds silence.

Language shapes how we think, debate, and understand the world. When our language is filtered through fear of algorithms rather than guided by shared norms and open discussion, internet freedom erodes, not with a ban, but with a shrug.

A Quieter Internet Isn’t a Better One

I don’t believe in a lawless internet, and I don’t support speech that harms others. However, I do believe that people should be able to discuss real issues openly without feeling like they’re navigating a linguistic obstacle course.

“Pew pew” and “unalived” may sound harmless or even funny sometimes. But they’re symptomatic of a much deeper issue that isn’t funny at all: an online environment where clarity is risky, and silence feels safer.

If we care about the future of online discourse, we should be asking why so many users feel the need to censor themselves in the first place. Because an internet where people are afraid to speak openly isn’t just moderated; it’s diminished.

And once online freedom is unalived, bringing it back won’t be easy