As a brief summary of the two previous articles onsocial networks: in art. 1/3 we looked at the business model and why leaks and scraping make “private” fragile; in art. 2/3 we reviewed scams, spear phishing, the dangers of LinkedIn, doxing, metadata and digital hygiene practices.
In this third and final article of the 3/3 series: we focus on algorithmic manipulation, synthetic users, social/psychological impact, cultural trends, regulation and a practical kit for your day-to-day life.
From bots and sock puppets to “synthetic users”.
We have gone from classic bots and sock puppets (fictitious accounts created to influence opinions or circumvent restrictions, such as in forums or RRSS) to AI-generated profiles with a face, voice and plausible behaviors.
Some platforms have hinted at introducing “AI users” with bios and profile pictures to generate and share content.
The line between human interactions and synthetic activity is blurred, degrading the social experience (who is talking? why?).
When engagement depends on artificial activity, trust is eroded and public debate becomes more manipulative.
Moderation, amplification and accountability
The digital public conversation is mediated by opaque algorithms that prioritize screen time and engagement engagement.
In many countries there is growing regulatory interest in distinguishing passivehosting from active algorithmic amplification.
To prevent platforms from abusing and manipulating us at the whim of anyone who puts a good sum of money on the table or forces them from unscrupulous governments, a sensible framework would require technical audits, log retention and proportionate duties of care, to avoid both impunity and over-censorship.
It is true that we also see some judicial systems, such as the European one, moving towards invoking greater responsibility of platforms for the content they allow in their space such as hate, racism or incitement to violence, demanding monitoring and removal.
If certain limits cannot be placed on what is said through them, the consequences can be disastrous for society, however much many call this censorship.
The regulatory pendulum is swinging; the key will be to balancefreedom of expression and harm reduction.
Geopolitics of feeds: the case of TikTok
The debate over banning TikTok in the U.S. is presented as a matter of national security and foreign manipulation.
The irony: for years, other powers have used their networks to influence outside their borders.
In any case, focusing only on one platform does not solve the systemic problem: business models based on tracking that feed optimized recommenders to engage and move opinion.
Without a cross-cutting approach, the result is arbitrary vetoes that standardize force solutions without attacking the root cause.
Why do you think the US is so worried about TikTok? Could it be because no one knows better than them what can be done with a social network with so many active users? Without a doubt, they are the perfect tools for the manipulation of society!
And don’t be fooled, when a platform like Facebook claims to be very secure, it must prove it, and the best way is to allow technical tests*(bug bounty*, audits, E2E encryption with public specifications, transparency reports with verifiable metrics).
On the other hand, we have the expansion of dating apps, which are bringing countless side effects: matching algorithms that can amplify inequalities (preference for homogeneity in education, income, beliefs), in addition to optimizing ephemeral encounters, far from helping to find a partner, they strive not to find one so that you can stay in them longer - better for business - over the welfare of the user.
It is the perfect example of how “optimizing” for platform metrics does not always maximize social value.
Also, I guess I don’t need to tell you that the vast majority of these dating apps are infected with scammers, spies waiting for the ideal victim, abandoned profiles that keep showing up for decades, synthetic profiles to give the feeling of having a large community, etc.
In the end, it can be said that all social networks are experts in psychological mechanisms (variable reinforcement, notifications, infinite scroll).
Increases in addiction, anxiety and depression linked to heavy use are continually being documented.
The promise of the metaverse - immersive realities that can be more engaging than everyday life - compounds the risk of social disconnection and isolation.
And don’t get me wrong, my intention, rather than demonizing technology (as was once done with television), is to recognize dangerous patterns and regulate with criteria focused on public health and transparency.
Movements are beginning to emerge, mainly in Gen Z, that are escaping from infinite scroll and influencer culture without substance, to spaces with community and human curation (blogs, newsletters, photoblogs).
It is not nostalgia: it is exhaustion in the face of algorithms that fragment conversations and reward noise. Returning to open protocols (RSS) and mailing lists is, for many, a relief.
Hints: how to detect algorithmic manipulation in your feed
- Improbable growth: new accounts with abrupt peaks of followers/likes and inhuman schedules.
- Coordinated choruses: identical or almost identical messages within seconds/minutes; chain responses that support a narrative.
- Misalignedengagement: many “likes” with few genuine comments; generic responses that do not refer to the content.
- Cardboard profiles: stock photos, vague bios, subtle localization/language errors; overly neat histories.
- Repeated “hook” topics: rotation of controversial issues with similar time patterns and hashtags.
- Synthetic signals: hyper-realistic avatars, cloned voices, “reactions” with slight desynchronies; responses that avoid verifiable details.
- Ephemeral links: newly registered domains, with no reputation, aggressive trackers and opaque privacy policies.
What you can do (actionable checklist)
- Data hygiene: limit permissions, disable advertising IDs, use site-isolated browsers and filter lists.
- Never upload documents or travel data in real time.
- 2FA on all your accounts; check active sessions frequently.
- Disable geolocation and clean metadata before sharing images.
- Do not share your address book with any app.
To conclude this series on social media, we must recognize that they are charming and engaging, but their business model and the ecosystem of manipulation require lucidity.
I’m not telling you to shut them down, but I am telling you to use them wisely, with the right tools and conscious habits, you can enjoy their advantages without becoming an easy target.