Surveillance capitalist companies study people to better manipulate their behaviour by advertisements. That skill can be used to attack a vigilant target by using oblivious mediators as a drone swarm delivering a payload.
The documentary “The Social Dilemma” (2020) pointed out ways to manipulate large numbers of users of smart devices, and how that leads to drastic changes in society. It claims that personalised nudges are being delivered to billions of people daily, with the apparent goal (or emerged effect) that the sum of all nudges leads to the subjugation of the whole society, while successfully avoiding detection by national security agencies and getting banned from countries.
Examples of that type of mass manipulation: Trump and Brexit were aided by personalised political propaganda on Facebook. Activists can’t improve the world while they’re busy fighting fascism. Brexit is a big distraction from matters important to Europe’s security right now.
Wikipedia: Facebook–Cambridge Analytica data scandal
In contrast, I seek ways to get through to specific vigilant targets that are avoiding big-tech devices, expecting to get surveilled and manipulated by them.
Which ways exist to manipulate the targets’ acquaintances to hamper the targets’ life goals? Especially political action opposing the technofascism that the proliferation of surveillance capitalism is leading to.
The resulting list of possible acquaintance swarm manipulations is meant as a warning for two kinds of people:
-
People thinking “My private information is worthless, I have nothing to hide.”
Wrong:
You enable the surveillance of people around you (hearing distance, wireless communication distance), some of which are trying to protect you and your country’s interests.
If you browse any media with a feed of recommendations, you enable the manipulation of what you know, think, and do.
Controlling the timing of notifications you receive on your smart device alters the timing of your other actions, which can be used to cause or prevent serendipitous meetings in real life, changing your future to the will of the manipulator, affecting the futures of people you interact with, like a chain of dominoes falling towards or away from vigilant targets your acquaintances interact with. -
People thinking “I was at risk, but I have stopped using any big tech devices or billionaire-owned media”.
Wrong:
You can be manipulated through the manipulated behaviour of your acquaintances. Small nudges can work in concert against you and your potential allies you don’t know yet, but whom the manipulator does know via surveillance, and is trying to prevent from gaining power.
The first group can be used against the second, unnoticed by national security agencies, to the detriment of both groups.
Which specific nudge swarm manipulations can currently be done to people and systems around targets to critically affect even the most vigilant targets?
Oblivious human swarm attacks:
-
Potential activists can be prevented from teaming up if they can be kept from finding each other. Their acquaintance networks can be kept insulated from each other by hiding certain recommendations on social media, and with other seemingly useful recommendations to the acquaintances, each activist can be strategically misdirected into separate small bubbles of activists below critical mass for effective action.
-
If a big-tech device of an acquaintance of an activist hears them mentioning (or having) symptoms of mild illness, any clues towards known cures can be hidden from their whole acquaintance network to prevent the activist from receiving the helpful information via hearsay, thereby prolonging the illness that is harming their activism.
-
If people around the people an activist interacts with in person are turned into antivaxxers, it increases the likelihood of the activist dying. Altering the timings of notifications and route plans in navigators cause close passes of people, forming a chain from a sick person to the target.
.
A question affecting the finesse of the manipulation is how much surveillance companies share data among each other. Do these companies sell each other enough user data to enable aligned, apparently coordinated manipulation of political enemies, or does each company have to work with their own data?
Some companies in the position to surveil and manipulate (and some of their products/services):
- Microsoft (Windows, Office, Bing, LinkedIn, Teams, Outlook, Hotmail, Copilot)
- OpenAI [27% Microsoft] (ChatGPT)
- Alphabet (Google, YouTube, Android, Chrome, Maps, Gmail, Gemini)
- Apple (iOS, Maps)
- Meta (WhatsApp, Facebook, Instagram, Threads)
- Amazon (Alexa, Prime, Kindle, Twitch)
- SpaceXAI (X/Twitter, Grok)
- Match Group (Match, Tinder, OkCupid, Hinge, Plenty of Fish, etc.)
- ByteDance (TikTok)
Signs of data sharing:
Online Behavioral Ads Fuel the Surveillance Industry—Here’s How
https://www.eff.org/deeplinks/2025/01/online-behavioral-ads-fuel-surveillance-industry-heres-how
U.S. judge rules Google must share search data with rivals, boosting Microsoft’s Bing
https://www.windowscentral.com/microsoft/google-ordered-to-help-rivals-like-microsoft-improve-bing-by-sharing-search-data-and-letting-android-phone-makers-preload-apps-like-edge-and-copilot
Apple, Google, Meta are sharing more data with the US government than ever, Proton finds
https://thenextweb.com/news/big-tech-sharing-more-data-with-us-government
Zuckerberg’s Meta considered sharing user data with China, whistleblower alleges
https://www.washingtonpost.com/technology/2025/03/09/meta-china-censorship-facebook-mark-zuckerberg/
Further reading:
(seemed relevant when I skimmed them)
Power in the modern ‘surveillance society’: From theory to methodology
https://doi.org/10.3233/IP-211536
Surveil to Manipulate in the Name of National Security: State Interference with Citizen Mental Sphere Via the Use of Information and Information Technologies
https://dx.doi.org/10.2139/ssrn.3888396
Facebook Shadow Profiles
https://ideas.repec.org/p/hal/wpaper/hal-04495641.html
The rise of authoritarian informationalism: escalating surveillance, manipulation, and control
https://www.tandfonline.com/doi/full/10.1080/13510347.2025.2579105
"
… shift from blunt repression towards more preventive, fine-grained strategies of control.
… enabling regimes to anticipate and pre-empt dissent, digital technologies reduce the need for overt coercion, allowing authoritarian rulers to maintain power while minimizing visible repression.
… more covert forms of control become harder to identify and counter.
"
Misdirection in Psychology: Unveiling the Art of Cognitive Deception
https://neurolaunch.com/misdirection-in-psychology/
"
Misdirection in psychology is the deliberate manipulation of attention to keep people from noticing something happening right in front of them, and it works because your brain simply can’t process everything your eyes take in. Magicians, marketers, and manipulators all exploit the same glitch: attention is a spotlight, not a floodlight, and whatever falls outside its beam might as well not exist.
"
False Narrative Psychology: Unraveling the Power of Deceptive Stories
https://neurolaunch.com/false-narrative-psychology/
"
False narrative psychology studies why our brains build, believe, and defend stories that aren’t true, whether about ourselves, other people, or entire events. The short answer: our minds aren’t built for accuracy. They’re built for coherence, and a coherent lie often feels more satisfying than a messy truth.
"
Watch:
The Social Dilemma (2020)
“The film argues that companies track and analyze users’ online activity in order to build artificial intelligence systems that predict and influence behavior.”
https://en.wikipedia.org/wiki/The_Social_Dilemma
https://tube.fediverse.at/videos/watch/812cee73-c4e7-4c23-9d1f-be577b4c290a (Official Trailer on PeerTube)
https://archive.org/details/the-social-dilemma-2020_202108 (The whole film)
Citizenfour (2014)
“Edward Snowden interviewed by journalists a week before public revelations about intelligence agencies’ mass surveillance programs all around the globe.”
https://en.wikipedia.org/wiki/Citizenfour
Watch on PeerTube:
https://tututu.tube/videos/watch/2b1612d1-2543-4673-8afc-a40fcf252ec1
Please leave the PeerTube tab open after watching, as it shares the video to other watchers, reducing server load.
I believe a good way to prevent covert strategic manipulation is the massive adoption of open-source products:
Linux and AOSP-derivative operating systems, Nextcloud, Signal messenger, Fediverse social media, etc. The Fediverse is unfortunately manipulable by bot accounts voting, commenting, and posting.
I was planning to feed the post to an AI chatbot, so I was looking for disambiguating words to prevent misunderstandings spawning sidetracks in its answer.
Why would you preach about the proliferation of subtle corporate manipulation, then subject yourself to the output of a biased text generator, most likely trained on manipulative material? What were you hoping to achieve that you couldn’t possibly have done better yourself?
As for the purpose, disambiguation through specific word choices may skew the LLM’s predictions to resemble other texts using those words, but if those words are used rarely in the training material, that might just feed back into association with more common, possibly ambiguous words instead, which would mean you’ve gained basically nothing. Even if it does produce sensible-looking output, there is no certainty that it will actually correlate with (let alone accurately reproduce) the content of relevant sources.
LLMs are not specialists. The idea of warding off “misunderstandings” is a mirage, because they don’t “understand” the subject in the first place. They know the way words appear together, like an undergrad memorising textbooks, and use that to assemble something that looks plausible and related to the prompt, but just as you shouldn’t trust a “passing grade is good enough” student with advanced topics, you shouldn’t rely on LLMs to handle such topics either.