# The Expanding Influence of Social Media Platforms on Society
Social media platforms have fundamentally transformed how information circulates, communities form, and power operates in contemporary society. What began as simple networking tools have evolved into sophisticated algorithmic ecosystems that shape political discourse, influence mental wellbeing, and redefine the boundaries between public and private spheres. With over 4.9 billion users worldwide spending an average of 2.5 hours daily on these platforms, the societal implications extend far beyond individual screen time. These digital infrastructures now mediate democratic processes, economic transactions, social movements, and personal identity formation at unprecedented scale and velocity.
The architecture of social media platforms isn’t neutral—it actively shapes what we see, how we think, and ultimately how societies function. From Facebook’s News Feed algorithms determining which information reaches billions of users, to Instagram’s design patterns triggering psychological responses, these systems operate according to principles that prioritise engagement and data extraction over public good. Understanding these mechanisms has become essential for navigating modern civic life, protecting mental health, and maintaining democratic integrity in an increasingly platformised world.
## Algorithmic Content Curation and Echo Chamber Formation on Facebook and Twitter
The invisible hand guiding your social media experience operates through sophisticated algorithmic systems that determine which content appears in your feed and in what order. These recommendation engines have fundamentally altered information distribution patterns, moving from chronological timelines to personalised curation based on predicted engagement probability. The shift represents one of the most consequential changes in how humans access information since the printing press, yet most users remain unaware of the selection processes shaping their digital realities.
### Feed Ranking Mechanisms: EdgeRank and Timeline Algorithm Evolution
Facebook’s original EdgeRank algorithm, introduced in 2009, pioneered the concept of relevance-based content curation at massive scale. The system evaluated three primary factors: affinity between user and content creator, the weight assigned to different content types, and time decay as posts aged. This seemingly straightforward formula concealed profound implications for information flow. Posts from friends you interacted with frequently received priority, photos outranked text updates, and recent content dominated feeds—creating a system that rewarded certain behaviours and content formats whilst marginalising others.
Modern iterations have grown exponentially more complex, incorporating thousands of signals including dwell time, comment sentiment analysis, video completion rates, and cross-platform behaviour patterns. Twitter’s algorithmic timeline, introduced in 2016 over user protests, similarly shifted from pure reverse-chronology to relevance ranking based on engagement predictions. These systems employ machine learning models trained on billions of user interactions, continuously refining their predictions about what will keep you scrolling. The result is a feedback loop where algorithmic predictions shape user behaviour, which generates new training data, further reinforcing existing patterns.
### Filter Bubble Psychology: Confirmation Bias Reinforcement Through Personalisation
The personalisation mechanisms embedded in social media platforms intersect powerfully with fundamental cognitive biases, particularly confirmation bias—our tendency to seek information that validates existing beliefs. When algorithms learn that you engage more with certain political perspectives or topical framings, they systematically increase your exposure to similar content whilst reducing visibility of contradictory viewpoints. This creates what internet activist Eli Pariser termed « filter bubbles »—personalised information ecosystems that isolate users within their own ideological universes.
Research demonstrates these effects empirically. A 2015 study published in Science found that Facebook users’ feeds contained approximately 70% ideologically concordant content, with the algorithmic curation reducing exposure to cross-cutting viewpoints by about 5% for conservatives and 8% for liberals. Whilst these percentages might seem modest, at population scale they represent billions of potential cross-ideological encounters that never occur. The cumulative effect over years of daily exposure shapes not just what we believe, but what we perceive as the range of acceptable opinion—the Overton window of our digital age.
### Engagement Metrics Manipulation: How Likes and Shares Shape Information Distribution
The fundamental business model of social media platforms rests on capturing and monetising attention, creating perverse incentives where engagement metrics become the primary determinant of content visibility. Posts that generate likes, shares, comments, and clicks receive algorithmic amplification regardless of their accuracy, nuance, or social value. This dynamic systematically privileges content that triggers emotional responses—particularly anger, outrage, and fear—over balanced, factual reporting that might be informative but less viscerally compelling.</p
Internal research at Meta has repeatedly shown that content eliciting strong emotional arousal—especially moral outrage—spreads more rapidly and receives disproportionate visibility. A widely cited 2018 study in Science analysing 10 years of Twitter data found that false news stories were 70% more likely to be retweeted than truths and spread to more people, faster. The mechanism is structurally simple: when we reward content with reactions, comments, and shares, we train both the algorithms and the humans producing content to prioritise whatever triggers those responses. Over time, creators learn to craft posts optimised not for accuracy or civic value, but for virality, and the algorithm dutifully amplifies the most provocative voices in the room.
For users, this engagement-driven sorting of information can feel like an unbiased reflection of “what everyone is talking about” when in reality it is closer to a popularity contest turbocharged by machine learning. Nuanced, context-rich reporting tends to lose out against sensationalised headlines and meme-ified talking points that demand instant emotional reactions. The danger is not just individual misperception; it is the gradual reshaping of public discourse around extremes, conflict, and hot takes. If you are not aware of how this invisible optimisation works, it is easy to mistake an engagement-optimised feed for an objective window onto reality.
### Political Polarisation Patterns: Cambridge Analytica Case Study and Microtargeting
These algorithmic dynamics have profound political implications, particularly when combined with granular data profiling and microtargeted advertising. The Cambridge Analytica scandal exposed how Facebook data harvested from up to 87 million users via a personality quiz app was repurposed to build psychographic profiles used in political campaigns. Rather than broadcasting the same message to all voters, campaigns could deliver thousands of tailored ad variants calibrated to individual personality traits, fears, and grievances—often in ways that were invisible to public scrutiny.
This form of political microtargeting interacts dangerously with echo chamber effects. When a campaign can show one group of voters a hard-line law-and-order message and another group a moderate, inclusive pitch—without either side seeing what the other receives—the possibility of shared democratic debate erodes. Studies of political advertising during both the 2016 Brexit referendum and the U.S. presidential election documented highly segmented, fear-based messaging about immigration, crime, and economic collapse that spread largely through Facebook’s ad infrastructure. The result is a political environment where citizens not only disagree on values but also inhabit different informational universes about basic facts and policy trade-offs.
Regulators have begun to scrutinise these practices, but clear rules are still emerging. Should there be limits on the granularity of political ad targeting—for example, banning targeting below a certain audience size or prohibiting the use of sensitive attributes such as race, religion, or inferred mental health? Should platforms be required to maintain public, searchable archives of all political ads and their targeting parameters? These questions go to the heart of how we balance free expression, commercial innovation, and the integrity of democratic processes in a platform-dominated information ecosystem.
Mental health implications of instagram and TikTok usage patterns
While Facebook and Twitter reshape political discourse, visually driven platforms like Instagram and TikTok exert their influence most powerfully on self-image, social comparison, and emotional wellbeing. Their interfaces are optimised for short-form, high-intensity content that demands little cognitive effort but delivers a steady drip of stimulation. For adolescents and young adults, who are still forming their identities and social hierarchies, the combination of constant connectivity, public metrics, and aesthetic ideals creates a psychological environment unlike anything previous generations have faced.
Dopamine-driven design: infinite scroll and variable reward schedules
Instagram and TikTok are built around what behavioural designers call “dopamine loops”—interaction patterns that trigger the brain’s reward circuitry in unpredictable ways. Features like infinite scroll, autoplay, and algorithmic “For You” recommendations remove natural stopping cues, turning what might have been a quick check-in into a 45-minute session. The core mechanism resembles a digital slot machine: every swipe or refresh is a small gamble that the next video, like, or notification will be especially rewarding.
Neuroscientific research shows that variable reward schedules—where positive reinforcement arrives at irregular, unpredictable intervals—are particularly effective at reinforcing behaviour. This is why casinos design slot machines the way they do, and why you sometimes find yourself pulling to refresh your feed without quite remembering when you opened the app. Over time, your brain learns to associate boredom or low mood with the urge to check social media for a quick hit of novelty or validation. For many users this remains manageable, but for a subset—especially those already vulnerable to anxiety or depression—the pattern can begin to resemble a behavioural addiction that displaces sleep, study, and offline relationships.
Body dysmorphia and aesthetic surgery trends: the instagram face phenomenon
One of the most visible cultural byproducts of Instagram has been the rise of what cosmetic surgeons call the “Instagram Face”: a hyper-filtered aesthetic defined by poreless skin, plumped lips, narrowed noses, and exaggerated eyes. Beauty filters and editing apps make these features appear effortlessly attainable, even though they often require extensive cosmetic procedures, expensive skincare, or simply digital manipulation. When your exposure to other people’s faces is dominated by these idealised images, your perception of what is “normal” subtly shifts.
Clinicians have reported an increase in cases of body dysmorphic disorder and “Snapchat dysmorphia,” where patients seek surgery to resemble their filtered selfies. A 2021 survey by the American Academy of Facial Plastic and Reconstructive Surgery found that 79% of surgeons reported patients referencing social media images when discussing procedures, a sharp increase over the previous decade. For young users—especially girls—this constant comparison to unrealistic standards can fuel dissatisfaction, disordered eating, and chronic self-criticism. The problem is not just any single image but the cumulative effect of thousands of subtly edited faces presented as candid snapshots of everyday life.
FOMO and social comparison theory: quantified Self-Worth through metrics
Beyond appearance, platforms like Instagram and TikTok quantify social value through visible metrics: follower counts, likes, views, and comments. Classic social comparison theory suggests that we evaluate ourselves in relation to others, especially peers. Social media transforms this innate tendency into a gamified leaderboard where popularity and status are publicly displayed and endlessly updated. Even if you intellectually know that people curate their highlight reels, it is hard not to feel left behind when your feed suggests that everyone else is more attractive, more successful, and having more fun.
This dynamic is closely linked to FOMO—the fear of missing out. When you see friends at gatherings you were not invited to, or creators your age achieving viral success, your brain interprets these as signals about your own social standing. Over time, you may start to measure your self-worth against engagement metrics rather than internal values or offline relationships. Some platforms have experimented with hiding like counts to reduce this pressure, but the underlying comparison mechanisms remain. The challenge for users is to develop conscious habits—such as time limits, intentional follows, and periodic digital detoxes—that prevent their sense of self from becoming fully tethered to fluctuating algorithmic feedback.
Teen depression correlations: frances haugen whistleblower revelations
The mental health impact on adolescents became a global headline in 2021 when Facebook whistleblower Frances Haugen leaked internal research showing that Instagram use was associated with increased rates of anxiety, depression, and body image issues among teenage girls. One internal slide reportedly stated that “We make body image issues worse for one in three teen girls,” and another highlighted that teens blamed Instagram for increases in the rates of anxiety and depression. While correlation does not prove causation, the documents suggested that Meta was aware of these risks yet moved cautiously on mitigation, in part because emotionally vulnerable users also tend to be highly engaged users.
Academic studies echo these concerns, though the effect sizes at the population level are often modest and heterogeneous. A 2022 review commissioned by the UK government concluded that there is a small but significant association between heavy social media use and poorer adolescent mental health, particularly for girls, but also stressed that the quality of existing research is uneven. Crucially, the impact appears to depend less on total time spent and more on what young people are doing online and how it intersects with offline vulnerabilities. This suggests that blanket screen time limits are a blunt tool; more targeted interventions might involve safer recommendation systems for minors, robust reporting and moderation of harassment, and tools that make it easier to curate supportive, interest-based communities.
Misinformation propagation mechanisms across platform architectures
Beyond individual wellbeing and political polarisation, one of the most contested impacts of social media is its role in amplifying misinformation. From election conspiracies to health hoaxes, false narratives can now travel across the globe in hours, morphing as they go. To understand why, we need to examine not only human psychology but also the network structures and technical affordances of different platforms. The way information spreads on Twitter or X is not identical to Facebook, WhatsApp, or TikTok, yet common patterns emerge when messages are optimised for speed and virality rather than verification.
Viral cascade dynamics: network theory and information diffusion models
Network theory provides a useful lens for understanding how misinformation ripples through digital platforms. Social networks are typically “scale-free,” meaning a small number of highly connected hubs—celebrities, influencers, large pages—have vastly more connections than the average user. When these hubs share a piece of content, it can trigger a viral cascade, propagating through successive layers of followers and their followers. Mathematical models of information diffusion show that the combination of high connectivity and low friction (one-click sharing, retweeting, forwarding) dramatically accelerates spread compared with older media ecosystems.
Crucially, misinformation often enjoys a structural advantage in these systems because it is unconstrained by facts and can be crafted for maximum emotional punch. When you combine a highly connected network with algorithms that privilege engagement, misleading but attention-grabbing posts are more likely to jump from small clusters into the wider graph. The result is a digital wildfire: once a narrative ignites in the right conditions, it becomes costly and time-consuming to extinguish, even if fact-checks eventually catch up. For policymakers and platform designers, this raises a difficult question: how do we preserve the benefits of rapid information sharing—such as during natural disasters—while reducing the systemic risk of viral falsehoods?
Bot armies and coordinated inauthentic behaviour detection systems
Not all participants in these information cascades are human. Automated accounts (“bots”) and coordinated inauthentic behaviour (CIB) networks can manufacture the illusion of consensus, amplify fringe narratives, and harass critics at scale. Research following the 2016 U.S. election and various European votes has documented bot networks pushing divisive hashtags, boosting polarising content, and strategically targeting journalists and activists. Similar tactics have appeared in commercial disinformation campaigns, where fake accounts promote products or smear competitors.
In response, platforms have developed increasingly sophisticated detection systems that blend machine learning with human investigation. These tools look for patterns such as high-volume posting, synchronised activity across accounts, shared IP addresses, and copy-pasted content. Facebook and Twitter now regularly announce takedowns of CIB networks linked to both state and non-state actors. Yet detection remains a cat-and-mouse game: as platforms refine their tools, adversaries adapt, using more human-operated “sock puppet” accounts, outsourcing operations to third-party firms, or shifting to encrypted and closed groups where monitoring is harder. For ordinary users, this means we cannot always assume that apparent grassroots enthusiasm or outrage is entirely organic.
COVID-19 infodemic: WHO classification and platform response strategies
The COVID-19 pandemic crystallised these dynamics into what the World Health Organization labelled an “infodemic”—an overabundance of information, some accurate and some not, that makes it hard for people to find trustworthy guidance. False claims about miracle cures, vaccine side effects, and conspiracy theories about the origins of the virus spread rapidly across Facebook, YouTube, WhatsApp, and newer platforms. In some countries, health authorities struggled to compete with viral memes and influencer videos that undermined public health advice.
Under intense pressure from governments and civil society, platforms rolled out a suite of response strategies. These included labelling or down-ranking misleading posts, removing egregious falsehoods, elevating information from trusted health organisations, and banning repeat offenders. WhatsApp limited message forwarding to slow chain messages, YouTube demonetised or removed harmful content, and TikTok introduced COVID-19 information hubs and in-app prompts. While these measures mitigated some harm, they also sparked debates about censorship, bias, and who gets to decide what counts as “authoritative” knowledge in fast-evolving situations.
Fact-checking infrastructure: Third-Party verification partnerships and limitations
One cornerstone of platform responses to misinformation has been the development of third-party fact-checking partnerships. Organisations accredited by bodies like the International Fact-Checking Network review viral posts, rate their accuracy, and provide contextual explanations. Platforms then use these ratings to apply labels, reduce reach, or in some cases remove content. In theory, this creates a feedback loop where high-quality corrections follow harmful claims and help users recalibrate their beliefs.
In practice, fact-checking faces structural limitations. Reviews are labour-intensive and always lag behind the speed of virality; by the time a claim is debunked, it may have reached millions. Corrections also rarely spread as far or as fast as the original falsehood, a phenomenon known as the “truth sandwich” problem. Moreover, in polarised environments, fact-checkers themselves can become targets of partisan suspicion, with their work dismissed as biased or politically motivated. For users, the most effective strategy is often a combination of healthy scepticism—pausing before you share, checking multiple sources—and using platform tools to report and flag dubious content rather than unintentionally boosting it.
Platform monopolisation and digital public sphere transformation
As social media platforms have grown, a handful of companies now mediate an enormous proportion of global digital communication. This concentration of power raises questions not only about competition and consumer choice but also about the health of the digital public sphere. When a small number of privately owned, advertising-funded platforms set the rules of online speech, they effectively govern a vast, transnational public square without the checks and balances that apply to states.
Meta’s ecosystem dominance: WhatsApp, instagram, and facebook integration
Meta (formerly Facebook, Inc.) illustrates this consolidation vividly. Through acquisitions of Instagram in 2012 and WhatsApp in 2014, the company built an integrated ecosystem that touches messaging, photo and video sharing, groups, and marketplace transactions. In many countries, WhatsApp has become the default communication channel; in others, Instagram and Facebook dominate social discovery and news consumption. Even when users believe they are diversifying by switching apps, they often remain within Meta’s broader data and advertising infrastructure.
The company’s ongoing efforts to integrate backend systems—such as unifying messaging infrastructure across Messenger, Instagram Direct, and WhatsApp—further blur the boundaries between services. From a user perspective, this can increase convenience: you can reach contacts across apps, share content seamlessly, and maintain a single identity. From a competition and privacy perspective, it raises significant concerns. Deep integration makes it harder to break the company up in antitrust remedies and concentrates more behavioural data in a single corporate entity. It also increases the systemic risk of outages or security breaches affecting billions of people at once, as seen in Meta’s major global downtime incidents.
Network effects and switching costs: why competition fails to emerge
One reason dominant platforms are so hard to dislodge is the strength of network effects. The value of a social network increases with each additional user: you join the platform where your friends, family, colleagues, and favourite creators already are. This creates a self-reinforcing cycle where “the winner takes most,” and smaller competitors struggle to achieve critical mass. Even when new entrants like TikTok succeed, they often do so by carving out a specific niche (short-form video, in this case) rather than replacing incumbents entirely.
Switching costs further entrench this dominance. Your social graph, message history, photos, and reputation are typically locked inside proprietary systems with limited portability. Leaving a platform may mean losing years of digital memories and social capital. While regulations like the EU’s Digital Markets Act push for greater data portability and interoperability—allowing users to move contacts and content between services—implementation is still in its early stages. Until meaningful interoperability becomes a reality, most users will remain effectively “captive” to the platforms where their communities already live.
Section 230 protection and content moderation liability debates
In the United States, debates about platform power are deeply intertwined with Section 230 of the Communications Decency Act, a 1996 law that shields online intermediaries from liability for most user-generated content while allowing them to moderate in “good faith.” Proponents argue that Section 230 enabled the growth of the modern internet by protecting platforms from being sued over every post, comment, or video. Critics counter that the law gives tech giants too much freedom to profit from harmful content without sufficient accountability.
Globally, variations on this liability debate are playing out as governments consider new regulations around hate speech, misinformation, and harmful but legal content. Some proposals seek to make platforms more responsible for rapid removal; others emphasise transparency, due process for users, and systemic risk assessments over individual takedown obligations. The core challenge is finding a balance where platforms cannot simply shrug off the societal consequences of their design choices, yet are not forced into heavy-handed censorship that undermines free expression. Whatever path regulators choose will profoundly shape the future of the digital public sphere.
Data monetisation models and surveillance capitalism economics
Underpinning many of these dynamics is the economic logic of surveillance capitalism—a term popularised by scholar Shoshana Zuboff to describe business models that turn human experience into behavioural data for profit. Social media platforms are “free” at the point of use because the product being sold is not the app but your attention and predictability. Every click, pause, and interaction feeds into models that try to forecast what you will do next, what you might buy, and which content is most likely to keep you engaged.
Behavioural surplus extraction: zuboff’s framework applied to social platforms
Zuboff’s framework distinguishes between data required to provide a service and what she calls “behavioural surplus”—extra information harvested to predict and shape future behaviour. On social media, this surplus goes far beyond basic account details or friend lists. It includes inferred traits such as political leanings, personality type, relationship status, mood, and even sleep patterns, derived from signals like posting times, language use, and interaction networks. These inferences are then bundled into advertising segments or fed into recommendation engines that aim not just to reflect your preferences but to nudge them.
When applied at scale, this extraction of behavioural surplus creates new asymmetries of knowledge and power. Platforms know vastly more about users than users know about how they are being profiled. This one-way mirror enables what some researchers call “behavioural modification markets,” where advertisers pay not merely for impressions but for the increased likelihood that you will take a desired action—click, sign up, vote, or even stay on the platform longer. The ethical concern is not targeted advertising per se, but the opacity and depth of the profiling, which users rarely understand or meaningfully consent to.
Attention economy metrics: cost per click and engagement rate optimisation
The translation of behavioural data into revenue happens through an array of advertising metrics and optimisation strategies. Cost per click (CPC), cost per mille (CPM, or thousand impressions), and cost per acquisition (CPA) quantify how much advertisers pay for different outcomes. Platforms use real-time auctions and machine learning to match ads with users who are statistically most likely to respond, constantly tweaking delivery to maximise engagement and revenue. In this attention economy, every design decision—from notification badges to autoplay videos—is evaluated in terms of its impact on key metrics like daily active users and session length.
For businesses, these tools can be extraordinarily powerful, enabling precise targeting and measurable returns. For society, the aggregated effect is an ecosystem where all incentives push toward capturing more attention with more emotionally charged, personalised content. If you are a marketer, this places a responsibility on you to consider not just short-term campaign performance but also the broader information environment you are helping to shape. Are your strategies contributing to meaningful engagement, or are they adding to the sea of distraction and misinformation that already overwhelms many feeds?
GDPR and data protection regulatory responses to privacy erosion
In response to growing concerns about privacy erosion, regulators—especially in Europe—have introduced comprehensive data protection frameworks. The EU’s General Data Protection Regulation (GDPR), which came into force in 2018, established stringent requirements for consent, data minimisation, user access, and breach notifications. It also empowered regulators to levy substantial fines, which have since been applied to several major tech companies for violations ranging from unlawful data processing to inadequate transparency.
GDPR and related laws, such as the California Consumer Privacy Act (CCPA), have begun to shift industry practices, prompting clearer privacy policies, consent pop-ups, and options to download or delete your data. Yet critics argue that the core surveillance advertising model remains largely intact, with most users clicking “accept all” simply to access services. Emerging proposals, including bans on certain types of behavioural advertising to minors and stricter limits on cross-site tracking, suggest that the next regulatory wave may target the economic foundations of surveillance capitalism more directly. For users and policymakers alike, the central question is whether meaningful privacy can coexist with business models that depend on ever-deeper data extraction.
Social movements mobilisation through platform affordances
Despite these risks, it is crucial to acknowledge that social media platforms have also become vital tools for social movements, enabling rapid mobilisation, storytelling, and global solidarity. The same features that allow misinformation to spread—low publishing barriers, instant sharing, and networked communication—also empower marginalised communities to document abuses, coordinate protests, and pressure institutions. The impact of these digital affordances on activism is complex: they can amplify voices and accelerate change, but they can also foster slacktivism, surveillance, and backlash.
Arab spring and twitter revolution narratives: critical analysis
The wave of uprisings across the Middle East and North Africa from 2010 onwards—often dubbed the “Arab Spring”—was the first major event widely framed as a “Twitter revolution.” Platforms like Facebook, Twitter, and YouTube played visible roles in organising demonstrations, sharing real-time updates, and broadcasting images of state violence to global audiences. Hashtags helped protesters coordinate meeting points, while citizen journalists uploaded footage that traditional media struggled to access under authoritarian regimes.
However, subsequent research has cautioned against attributing these complex political events primarily to social media. Digital tools were accelerants rather than root causes, and their impact varied across countries with differing levels of internet penetration, literacy, and state repression. Moreover, the same platforms that enabled mobilisation also facilitated surveillance and counter-mobilisation by security services and rival groups. The Arab Spring illustrates both the promise and the peril of networked activism: social media can lower the cost of collective action and make regimes more visible, but it does not substitute for organisational capacity, institutional reform, or long-term political strategy.
Black lives matter hashtag activism and digital protest organisation
The Black Lives Matter (BLM) movement offers a more recent example of sustained digital activism. Originating as a hashtag in 2013 after the acquittal of George Zimmerman in the killing of Trayvon Martin, #BlackLivesMatter evolved into a decentralised network of local chapters and supporters. Social media platforms became essential for documenting police violence, coordinating protests, and circulating educational resources on systemic racism. During the global demonstrations following the murder of George Floyd in 2020, videos and livestreams shared on Instagram, Twitter, Facebook, and TikTok galvanised millions worldwide.
At the same time, BLM’s digital presence highlighted tensions within online activism. Hashtag participation can create the illusion of engagement without deeper commitment—what critics term “performative allyship.” Algorithms sometimes prioritised sensational or confrontational content over nuanced discussions of policy demands such as qualified immunity reform or community investment. Yet, despite these challenges, the movement demonstrates how social media can keep issues in the public eye, pressure brands and institutions to respond, and provide marginalised communities with tools to narrate their own experiences rather than relying on legacy media filters.
Telegram and signal migration: encrypted communication for civil resistance
As awareness of platform surveillance and moderation has grown, many activists have migrated parts of their organising infrastructure to encrypted messaging apps like Telegram and Signal. These tools offer end-to-end encryption, self-destructing messages, and, in Telegram’s case, large group channels that can broadcast updates to hundreds of thousands of subscribers. From pro-democracy protests in Hong Kong and Belarus to climate activism and labour organising, encrypted channels have become crucial for coordination under conditions of risk.
However, this shift also presents new challenges. Encryption protects legitimate dissenters and vulnerable communities, but it can equally shield extremists, criminals, and disinformation networks from oversight. Governments have responded with attempts to weaken encryption, mandate backdoors, or compel companies to provide metadata and access under certain conditions—moves that civil liberties groups warn could undermine privacy for everyone. For movements using these tools, the strategic question is how to balance security with reach: highly secure channels may be excellent for internal coordination but poor for mass outreach, whereas public platforms offer visibility at the cost of greater exposure to surveillance and content takedowns.