Featured

Who Checks the Checker?

Petru Cojocaru Google 08 December 2025 Hits: 251 29 minutes read

When the algorithm becomes the censor and people forget they are allowed to ask questions

My father never knew how to use a search engine. He repaired engines — engines from Dacias, AROs, agricultural machinery that came to him from all around Botoșani, each with a wrong sound or a wrong silence. He could tell from the sound alone where the problem lay. He needed no manual, no computer. He needed attention. And years of accumulated attention.

I think of him now, when I open Google and see that the search engine I have used for twenty years has changed. Not radically, not visibly — but subtly, insidiously. A layer has grown over the results. A voice that belongs to no one. An answer before the question.

Google calls it AI Overview. I call it something else. I call it the problem of who checks the checker — and it took me a long time to understand why it disturbs me so profoundly.

1. Neutrality as a Convenient Lie

In official communications — calibrated by entire departments of well-paid people trained to sound natural — Google obsessively insists on a single word: neutrality. Our algorithms are neutral. Objective. Data-driven.

It is a beautifully packaged lie.

Every algorithm is built by people. Every line of code reflects choices: which criteria matter, which sources take priority, which answers are 'correct'. There is no such thing as technological neutrality. There is power disguised in code. There are people on five-figure monthly salaries, with specific beliefs about the world, deciding what the planet sees over breakfast.

My father never talked about neutrality when he repaired engines. He talked about what worked and what did not. The difference was clear. Audible. Verifiable. If he made a mistake, the car did not start. Accountability was immediate, concrete, non-negotiable.

Google does not have that accountability. It can err in silence. It can filter in silence. It can omit in silence. And no one will ever know precisely what was omitted — because there is no sound of the engine that will not start.

That is the first and most important lie built into the system: the idea that technology can be neutral. Those who believe it are naïve. Those who preach it from offices overlooking the Pacific are neither naïve nor innocent — they are the architects of systemic manipulation.

2. The Invisible Mechanism: How Filtering Becomes Censorship

When Google AI generates a response to your query, the process is roughly as follows — partially documented in technical papers, the remainder reconstructed by independent researchers through reverse engineering:

Stage 1 — The algorithm selects 20–50 sources from its index. Stage 2 — The algorithm evaluates the 'quality' and 'credibility' of each source. Stage 3 — The algorithm prioritises sources according to its programmed criteria. Stage 4 — The AI synthesises information from the selected sources. Stage 5 — The final answer appears on your screen.

The fundamental problem lies in Stages 2 and 3. Who defines 'quality'? Who defines 'credibility'? On the basis of what assumptions? With what implicit agendas?

Google systematically refuses transparency. The official reason: protecting the algorithms from manipulation. The real translation: we do not want you to see how we filter information according to our own worldview.

The Electronic Frontier Foundation documented in 2024 what happens when you test the same query on contentious subjects — climate change, vaccine efficacy, geopolitical conflicts. AI Overview systematically prioritised mainstream sources whilst marginalising or entirely eliminating dissenting voices — even when those dissenting voices came from academia and held solid credentials.

We are not talking about the removal of extremist propaganda. We are talking about the systematic marginalisation of any perspective that challenges the dominant consensus, regardless of the quality of the arguments.

The thought that haunts me: if my father had worked in a garage with opaque procedures — if no one had been able to verify how he diagnosed an engine, which parts he replaced, why he chose one solution over another — would we still have gone to him? With the same trust?

No. We would have checked up on him. We would have held him to account. That is the fundamental right of anyone who pays for a service.

No one holds the algorithm to account.

3. Biases Amplified on a Planetary Scale

Every person has prejudices. They are inevitable, they are human, they are part of the texture of existence. The problem arises when the prejudices of a small group are technologically amplified on a global scale.

Google AI is trained on data from the internet. The internet is dominated by: English — over 60% of content; Western, urban, university-educated perspectives; mainstream corporate media and establishment academic institutions.

The result is predictable: the AI reflects the dominant biases of the dominant sources. This is not neutrality. It is cultural hegemonism dressed up in a friendly interface.

UNESCO researchers documented in 2024 what happens when you ask about traditional medical practices. AI Overview exclusively prioritised modern Western medicine and described African or Asian traditional practices in terms of 'folk beliefs' or 'lacking scientific basis' — even when pharmacological and anthropological studies demonstrated their efficacy for specific conditions.

I tested this myself, in my own modest way, within the Romanian context. Queries about Romanian history, traditions, and cultural values. The responses synthesised predominantly from English-language sources, automatically translated — external perspectives about Romania, not authentic Romanian voices. When I asked 'what are the fundamental values of Romanian culture?', I received something generic and superficial, akin to what a Western tourist might say after a fortnight in Bucharest. No depth. No nuance. No understanding from within.

Jean Baudrillard spoke of 'simulacral imperialism' — the moment when the Western representation of reality becomes more 'real' than the locally lived reality. Google AI materialises this imperialism with an efficiency that Baudrillard, in 1981, could never have imagined possible.

Noica wrote somewhere that a nation is not a geographical accident but a form of spirit. As long as that spirit is not properly indexed, it simply does not exist for the algorithm that decides what is 'credible'.

4. Uniformisation: When Diversity Becomes a Monolith

An educator from Cluj told me something that stopped me in my tracks. His students, from different years and different backgrounds, had all been given the same assignment: Analyse the impact of globalisation. Every one of them had used Google AI Overview as their primary source. The essays were almost identical.

Not textually — the AI had personalised the wording. But conceptually, structurally, perspectivally: the same arguments, the same order, the same conclusions.

Intellectual diversity, artificially homogenised.

The problem is not that the AI provides correct or incorrect information. The problem is that it offers a single narrative, presented as Complete Truth. Alternative perspectives, nuanced interpretations, legitimate academic controversies — all disappear in the process of 'simplification'.

Plato built philosophy through dialogue. Socrates versus interlocutors, contradictory arguments, synthesis emerging from tension. Truth was not revealed — it was constructed collaboratively, through productive intellectual conflict. Google AI eliminates dialogue. It offers authoritarian monologue. There is no contradiction. There is no tension.

Hannah Arendt warned in The Origins of Totalitarianism: authoritarian systems do not need to actively persuade the masses. They need only to eliminate the capacity for independent thought. When people cease to think critically, they accept authority without resistance.

Google AI achieves technologically what classical totalitarianisms achieved through crude propaganda. The difference is that classical totalitarianisms were recognised as such. The algorithm is perceived as a neutral service.

My son uses these tools daily. I think about that. I think about what it means to grow up with an oracle that answers before you have truly formulated the question. What capacity do you have, at twenty, to live with uncertainty? To say I don't know? To sit with a question without resolving it immediately?

Uncertainty is the condition of authentic thought. The oracle eliminates it.

5. Sophisticated Disinformation: When Truth Becomes Negotiable

AI Overview presents all information with the same apparent authority. The format is identical. The tone is identical. The structure is identical.

The user cannot distinguish — from the presentation alone — between information grounded in solid scientific consensus, supported by thousands of studies, and information extracted from a dubious but well-indexed source. Everything looks equally 'true' because the formatting is identical.

Researchers at MIT documented in 2024 a technique called flooding: malicious actors — authoritarian governments, unscrupulous corporations, extremist groups — publish masses of coordinated content across diverse platforms. Content that appears credible: professionally designed websites, fictitious authors with elaborate biographies, 'studies' with fabricated methodology.

Google indexes everything. It does not automatically distinguish between legitimate research and sophisticated fabrication. When the AI synthesises from thirty sources for a query, five or ten of them may form part of a coordinated disinformation campaign.

The result: the final response is contaminated. Not completely false — close enough to the truth to appear credible, sufficiently skewed to distort understanding.

Truth becomes negotiable through volume. Whoever publishes more coordinated content influences what the AI synthesises.

Jaron Lanier warned: 'When reality becomes subject to algorithmic negotiation, civilisation collapses.' This is not dramatism. It is a sober observation grounded in an understanding of the mechanisms at work.

6. The Epistemological Monopoly: A Single Gatekeeper for Global Knowledge

StatCounter documents that Google controls over 92% of the global search market. In Romania — and in the majority of European countries — the figure exceeds 95%.

The direct translation: nine out of ten people access information through Google's filter.

A single corporate entity — democratically unelected, unaccountable to citizens, primarily motivated by profit — becomes the arbiter of truth for five billion people. No historical empire — Roman, Mongol, or British — simultaneously controlled what so many people believed about reality.

Orwell described in Nineteen Eighty-Four the centralised control of information. The crucial difference from 2025: in Orwell's dystopia, control was evident, brutal, acknowledged. Big Brother barked commands through compulsory television sets. Today, Big Brother whispers suggestions through the telephones we carry voluntarily, that we love, that we regard as extensions of our own minds.

Philip K. Dick explored in The Man in the High Castle what would happen if a single entity controlled what people believed about history. The answer: reality becomes malleable. Google now possesses that power — not in a fictional dystopia, but in contemporary reality.

And the fragility is systemic. When contemporary civilisation becomes progressively dependent on a single system for access to knowledge, that system becomes an existential vulnerability. What happens if it is compromised by malicious actors? If it malfunctions at scale? If it is wielded as a weapon by an authoritarian government?

The engineers of critical systems know this: excessive concentration produces catastrophic fragility. My father knew, in his own way, the same thing — you don't put all the power on a single cylinder.

7. Resistance and Its Illusions

The defenders of the status quo argue: if you don't like Google, use alternatives — DuckDuckGo, Bing, Brave.

It is an argument that sounds logical and is almost entirely divorced from reality.

Google controls over 92% of the global market. Bing holds 3%. DuckDuckGo holds 0.5%. And, more importantly: most 'alternatives' use Google's technical infrastructure — indexing, core algorithms, even results. They are not genuine alternatives. They are façades of diversity on a monopolistic foundation.

Cal Newport documents the phenomenon of technological lock-in — the moment at which a system becomes so dominant that migration becomes practically impossible. Google passed that threshold long ago.

And regulation? The European Union adopted the Digital Markets Act in 2023, designating Google a 'gatekeeper'. The reality of enforcement: token fines for a corporation with annual profits in the tens of billions, formal compliance with no substantive change in behaviour.

The Electronic Frontier Foundation explicitly documents the regulatory capture of overseers by the regulated industry. When corporations spend billions on lobbying and employ hundreds of legal experts, whilst regulatory agencies operate on limited budgets, the result is predictable: the regulations are effectively written by those being regulated.

The law becomes theatre. A public spectacle of 'regulation' that soothes citizens' anxieties without fundamentally altering the corporation's power.

8. The Moral Dimension: Discernment as Responsibility

The patristic writings — Saint John Chrysostom, Saint Maximus the Confessor, Abba Dorotheus — insist upon a concept that the modern world has hastily forgotten: discernment. The capacity to distinguish truth from falsehood, substance from appearance, genuine good from simulated good.

Discernment does not come automatically. It is formed through practice, through effort, through years of cultivated attention.

Google AI eliminates the need for discernment. It offers Truth ready-made. You no longer need to discern — you need only to accept.

From an Eastern Orthodox Christian perspective — fundamental to the cultural context in which I was raised and in which my son is growing up — this constitutes an abdication of moral responsibility. Man is created with free will and with reason precisely in order to discern. When he transfers that responsibility to an algorithm, he surrenders something essential from his humanity.

Abba Dorotheus taught that excessive dependence on external authority — even legitimate authority — atrophies one's own capacity for judgement. Spiritual maturity requires responsible autonomy. The principle applies perfectly to dependence on Google AI: constant use without critical thought produces moral and intellectual atrophy.

I am not proselytising. I observe that the traditional spiritual dimension converges, on this point, perfectly with secular concerns about autonomy, freedom, and civic responsibility.

Plato described in the Republic the process of the philosopher who leaves the Cave, sees reality directly, and then returns to help the others escape. It is an image of responsibility — the responsibility of the one who has seen towards those who have not yet seen.

This is not a grandiose mission. It is everyday responsibility: verify your sources, compare perspectives, resist the temptation of the instant answer, educate your children for critical thinking.

Every individual choice contributes to resistance or to collective capitulation.

9. The Community of Those Who Dared to Ask: A Portrait of Google's Critics

There is a category of people you will not find in AI Overview when you search for 'Google critics'. Or, if you find them, you will find them neutralised — presented as eccentric fringe figures, conspiracy theorists, nostalgic Luddites who have failed to accept progress. The algorithm knows how to discredit through labelling. It is simpler than engaging with arguments.

I know them differently. I have read them. Some of them have changed the way I look at a screen.

They are not a movement. They have no headquarters, no manifesto, no leader. They are people from different fields, with different vocabularies, who have arrived at the same questions by different routes. What unites them is not hatred of technology — it is love of something that uncontrolled technology threatens: free thought, epistemic diversity, individual responsibility.

Shoshana Zuboff — professor at Harvard Business School, author of The Age of Surveillance Capitalism — spent ten years building the argument that Google is not a search company. It is an extraction company. It extracts human behaviour as raw material, processes it into predictions, and sells those predictions. The user is not the customer. The user is the resource. Zuboff did not speak from the margins — she spoke from within the elite academy. She was systematically ignored by the mainstream tech press until she could no longer be ignored.

Jaron Lanier — one of the pioneers of virtual reality, a man who genuinely contributed to building the internet — wrote ten arguments for why you should delete your social media accounts. Not out of reaction, but out of a deep understanding of the mechanisms he helped to build. There is a particular kind of credibility in the voice of someone who says: I participated in constructing this system and I am now telling you what mistakes we made.

Eli Pariser coined, in 2011, the concept of the filter bubble. He demonstrated, with data, that personalisation algorithms do not show us a more relevant world. They show us a smaller world, a more comfortable world, a less contradictory world. A world from which uncomfortable perspectives have quietly vanished, without us noticing their disappearance.

The Center for Humane Technology — founded by Tristan Harris, a former design ethicist at Google — systematically documents how the design of tech platforms exploits human cognitive vulnerabilities. Harris worked from the inside. He saw how human attention is treated as a resource to be captured, not a value to be respected. He left. He spoke out.

There are also European voices, less visible in an algorithm that favours the Anglophone world. Evgeny Morozov, the Belarusian researcher based in Europe, critiques what he calls technological solutionism — the belief that every human problem has a digital solution, that efficiency is an end in itself, that optimisation is a moral good. It is a more subtle critique than the rejection of technology. It is a critique of the ideology that accompanies it.

And there are, of course, anonymous researchers — those who publish in academic journals read by no one outside their own field, who patiently and methodologically document precisely what I describe in this article: systematic filtering, structural biases, vulnerabilities to coordinated disinformation. They are invisible in AI Overview precisely because their work is the most dangerous to the system: it is verifiable, reproducible, and difficult to label away.

What do all these people have in common, beyond their specific arguments? A quality I recognise immediately, because I saw it in my father and in the people of the Botoșani of my childhood who knew their trade: the patience to understand before judging. They do not take a position before they understand the mechanisms. They do not condemn technology wholesale. They condemn its use in the service of concentrating power and atrophying humanity.

That is an important distinction — and one the algorithm deliberately levels. It is simpler to present a critique of Google as an irrational rejection of progress than to engage with the concrete arguments of Zuboff or Harris. The first option requires no answer. The second would require transparency.

I sometimes think that my son will grow up in a world where this community of critics — if it survives as an audible voice — will matter more than any regulatory institution. Not because institutions are unimportant, but because the real power to change the way people relate to a tool comes from understanding, not from law. The law can compel transparency. It cannot compel critical thinking.

That must be cultivated differently. Through books. Through conversations. Through the example of people who chose to keep asking, even when the answer was already available on the screen.

The community of Google's critics is not a club of the discontented. It is, in the most serious sense of the word, a community of those who still believe that the question is worth more than the answer.

Conclusion: Information Integrity at a Crossroads

The risks are not dystopian speculation. They are realities documented through independent academic research, reports from digital rights organisations, philosophical and sociological analysis, and a traditional spiritual doctrine that spoke of these very mechanisms — under different names — long before the internet existed.

The convergence is complete: the extreme concentration of epistemological power, combined with algorithmic opacity and corporate profit motives, produces a systemic vulnerability for global informational integrity.

Separating AI as a distinct module — transparent, optional, complementary to classical search rather than compulsorily overlaid upon it — would preserve the benefits of the technology whilst eliminating the concentrated risks. Google has the technical capability. It does not have the economic incentive.

The fundamental question transcends technology: what kind of world do we want to build?

One in which knowledge is centralised, filtered, and controlled by a private corporation motivated by profit? Or one in which information remains diverse, directly accessible, and subject to individual critical evaluation?

The first option produces maximum efficiency and total dependence. The second produces sustained effort and genuine autonomy.

History demonstrates repeatedly: societies that choose comfort over freedom lose both.

For the moment, the test suggests we are failing. The majority use these tools passively, without questions, without resistance. But stories change. When enough people understand the mechanisms — when enough parents see their children cognitively atrophied, when enough journalists realise they are being harvested without attribution, when enough researchers perceive the degradation of the academic ecosystem — resistance becomes possible.

My father did not check engines with a manual in front of him. He listened to them. He knew that an engine producing the wrong sound had something to say. That you had to pay attention. That the authority of diagnosis comes from years of presence, not from a certificate on the wall.

That kind of knowledge cannot be indexed. It cannot be synthesised. It cannot be delivered in a few seconds as a response to a query.

And for precisely that reason, it is worth defending.

The first step: understanding. The second: refusing automatism. The third: the conscious alternative.

The choice remains free. For now.

Questions and Answers

On the Mechanisms of Informational Control

1. Is Google AI Overview a neutral instrument for organising information? No. Algorithmic neutrality is a convenient fiction. Every algorithm is built by people who make implicit choices — which sources matter, which answers are 'correct', which perspectives are prioritised. There is no technological neutrality; there is only power disguised in code.

2. How does the process by which Google AI filters information before displaying it actually work? The algorithm selects 20–50 sources, evaluates them according to undisclosed criteria of 'quality' and 'credibility', prioritises them according to these opaque criteria, synthesises the response, and displays it to the user — without the user seeing what was omitted or why.

3. Are academic voices that challenge the dominant consensus marginalised in AI Overview results? Yes. The Electronic Frontier Foundation documented in 2024 that AI Overview systematically marginalised dissenting perspectives — including those from credentialed academic institutions — on topics such as climate change, vaccines, and geopolitical conflicts, exclusively prioritising mainstream sources.

4. Why is there no 'sound of the engine that won't start' in the case of Google's algorithmic errors? Because, unlike a mechanic whose mistake is immediately visible — the car does not start — the algorithm can err, filter, or omit in silence. The user receives no signal that information is absent. There is no feedback loop for failure.

5. What is it about the structure of the internet that makes Western bias inevitable in AI training? The internet is asymmetrically constructed: over 60% of content is in English, the dominant perspectives are Western, urban, and university-educated, and the prioritised sources are mainstream corporate media. An AI trained on this data cannot produce anything other than a faithful reflection of these structural inequalities.

6. Does Google AI reflect the cultural perspectives of the entire world, or predominantly Western ones? Predominantly Western. Non-dominant cultures — including Romanian culture — are represented through automatically translated external sources rather than authentic internal voices. Questions about Romanian values or history yield responses constructed from the perspective of an outside observer.

7. What happens to the critical thinking of young people who use Google AI daily as their primary source? It atrophies. When the AI offers a definitive answer before a question has been truly formulated, it eliminates the exercise of uncertainty — the fundamental condition of authentic thought. Students from different backgrounds who use AI Overview as their primary source produce conceptually identical essays.

8. Can sophisticated disinformation contaminate responses generated by Google AI? Yes. MIT researchers documented the 'flooding' technique: malicious actors publish masses of coordinated, apparently credible content. Google indexes everything, without automatically distinguishing between legitimate research and sophisticated fabrication. When the AI synthesises from dozens of sources, some of them contaminated distort the final response in a calculated direction.

9. Are there genuine alternatives to Google for internet searches? Practically speaking, no. Google controls over 92% of the global search market. Alternatives — Bing (3%), DuckDuckGo (0.5%) — are marginal, and most use Google's technical infrastructure. They are not genuine alternatives; they are façades of diversity on a monopolistic foundation.

10. What does the Orthodox patristic tradition say about dependence on external authority for judgement? Abba Dorotheus taught that excessive dependence on external authority — even legitimate authority — atrophies one's own capacity for judgement. Spiritual maturity requires responsible autonomy. Applied to the digital context: constant use of AI without critical thought produces moral and intellectual atrophy.

11. Is European regulation — the Digital Markets Act, the AI Act — sufficient to control Google's power? No. Although Google was officially designated a 'gatekeeper' under the Digital Markets Act (2023), enforcement in practice amounts to token fines and formal compliance. The Electronic Frontier Foundation explicitly documents the regulatory capture of overseers by the regulated industry: the regulations are effectively written by those being regulated.

12. What technical solution might preserve the benefits of AI in search whilst eliminating the risks of epistemological control? Separating AI as a distinct module — transparent, optional, complementary to classical search rather than compulsorily overlaid upon it. Google has the technical capability for this separation. It does not have the economic incentive to implement it.

13. How does Google AI Overview transform democratic debate from a necessity into a redundant act? Democracy functions on a plurality of perspectives and on public debate. When citizens believe that every question already has a definitive answer furnished by an authoritative algorithm, the need to listen to contradictory arguments disappears. Debate becomes superfluous, and the cognitive substrate of democracy erodes.

14. What is the 'flooding' technique and why does it make AI vulnerable to it? It is a technique documented by MIT whereby malicious actors publish masses of coordinated content with an appearance of credibility — professionally designed websites, fictitious authors with elaborate biographies, studies with fabricated methodology. Google indexes indiscriminately; when the AI synthesises from dozens of sources, a few contaminated ones skew the final response in a calculated direction.

15. Why cannot knowledge accumulated through direct experience be indexed or replaced by AI? Practical knowledge — the mechanic's diagnosis by sound, the judgement formed through years of presence — is not stored in text and cannot be compressed into trainable data. Its authority derives from continuity and personal accountability, not from the volume of information. The algorithm can replicate neither.

16. Through what concrete mechanism do minority cultures become invisible in Google AI responses? The AI synthesises predominantly from the most heavily indexed sources — which are overwhelmingly Anglophone and Western. For queries about non-dominant cultures, responses are constructed from automatically translated external perspectives rather than authentic internal voices. Romanian culture appears filtered through the eyes of an outside observer, not expressed from within.

17. What is the connection between Google's informational monopoly and the concept of 'single point of failure' in systems engineering? In critical systems engineering, dependence on a single node produces existential vulnerability. Google has become the central node through which contemporary civilisation accesses knowledge. Its compromise — through hacking, large-scale malfunction, or political pressure — would become a global informational catastrophe.

18. How does Baudrillard describe the phenomenon that Google AI materialises in relation to non-Western cultures? Baudrillard called it 'simulacral imperialism' — the moment when the Western representation of a reality becomes more 'real' than the locally lived reality. Google AI materialises precisely this mechanism: for billions of users, the external description of a culture becomes more accessible and more authoritative than that culture's own account of itself.

19. What would need to happen for collective resistance to Google's epistemological monopoly to become possible? The article identifies a threshold: when enough parents observe the cognitive atrophy of their children, when enough journalists realise they are being harvested without attribution, when enough researchers perceive the degradation of the academic ecosystem — the critical mass of understanding generates the pressure necessary for change.

20. What is the moral difference between surrendering discernment to an algorithm and surrendering it to any other external authority? According to the patristic tradition invoked in the article — Saint Maximus the Confessor, Abba Dorotheus — there is no essential difference: any transfer of one's own judgement to an external authority atrophies one's moral and intellectual capacity. What is specific to the algorithm is that the transfer occurs invisibly, without any awareness of the abdication.

On the Community of Those Who Critique Google

21. Who are the principal critics of Google's informational power, and what lends them credibility? They are not marginal figures or conspiracy theorists but people with solid biographies within the very system they critique: Shoshana Zuboff — professor at Harvard Business School, author of ten years of documented research; Jaron Lanier — pioneer of virtual reality and co-builder of the early internet; Tristan Harris — former design ethicist employed by Google; Eli Pariser — the researcher who conceptualised the filter bubble; Evgeny Morozov — analyst of technological solutionism. Their credibility comes from the inside, not the outside.

22. What unites Google's critics beyond the specific arguments of each? The patience to understand before judging. They do not condemn technology wholesale, nor do they reject it from pre-digital nostalgia. They condemn its use in the service of concentrating power and atrophying humanity — a distinction the algorithm deliberately levels by presenting any criticism as an irrational rejection of progress.

23. Why is criticism of Google systematically less visible in AI Overview results than its defence? The mechanism is circular: the algorithm prioritises sources it considers credible, and credibility is defined largely by alignment with the dominant consensus — which Google itself constructs. Critical voices, regardless of the solidity of their arguments, are marginalised through labelling or by simple absence from syntheses. It is simpler to ignore them than to answer them.

24. What does Evgeny Morozov specifically contribute to the critique of the Google ecosystem that the other cited authors do not? Morozov critiques not merely the corporation's behaviour but the ideology that legitimises it: technological solutionism — the belief that every human problem has a digital solution, that efficiency is a moral end in itself, that optimisation replaces judgement. It is a deeper critique than the behavioural one — it attacks the premises, not the consequences.

25. Why might this community of critics prove more important than regulatory institutions in changing our relationship with Google? The law can compel formal transparency — it cannot compel critical thinking. The real power to change how billions of people relate to a tool comes from understanding, not from a fine. The community of Google's critics cultivates precisely the understanding that the algorithm discourages: that the question is worth more than the answer instantly available on screen.

Online Bibliography

Risks to Information Integrity — Who Checks the Checker?

Organisations and Institutional Reports

Electronic Frontier Foundation — non-profit organisation for digital rights, systematically documenting abuses of the informational monopoly and the regulatory capture of overseers by industry.

Main page: https://www.eff.org
Competition and tech monopoly section: https://www.eff.org/issues/competition
Report: Court Orders Google (a Monopolist) To Knock It Off With the Monopoly Stuff (October 2024): https://www.eff.org/deeplinks/2024/10/court-orders-google-monopolist-knock-it-monopoly-stuff

Center for Humane Technology — non-profit organisation founded by Tristan Harris (former Google design ethicist), documenting the manipulative design of tech platforms and the exploitation of cognitive vulnerabilities.

Main page: https://www.humanetech.com
Mission and impact: https://www.humanetech.com/impact-and-story
Podcast Your Undivided Attention: https://www.humanetech.com/podcast

StatCounter Global Stats — real-time data on the global market share of search engines.

Search engine market share (global): https://gs.statcounter.com/search-engine-market-share
Desktop + mobile data: https://gs.statcounter.com/search-engine-market-share/desktop-mobile/worldwide

European Regulation — Digital Markets Act (2023)

Official EUR-Lex text: https://eur-lex.europa.eu/legal-content/RO/TXT/?uri=CELEX%3A32022R1925

European Regulation — AI Act (2024)

Official EUR-Lex text: https://eur-lex.europa.eu/legal-content/RO/TXT/?uri=CELEX%3A32024R1689

Books — Publisher Pages and Verified Information

Shoshana Zuboff — The Age of Surveillance Capitalism (2019) The central argument: Google is not a search company but an extraction company — human behaviour is raw material, the user is not the customer but the resource.

Author's page (Harvard Business School): https://www.hbs.edu/faculty/Pages/item.aspx?num=56791
Official book site: https://shoshanazuboff.com/book/
Publisher page (PublicAffairs / Hachette): https://www.hachettebookgroup.com/titles/shoshana-zuboff/the-age-of-surveillance-capitalism/9781610395694/
Wikipedia (EN): https://en.wikipedia.org/wiki/The_Age_of_Surveillance_Capitalism

Jaron Lanier — Ten Arguments for Deleting Your Social Media Accounts Right Now (2018) A pioneer of virtual reality and co-builder of the early internet argues that tech platforms are engines of behaviour modification, not communication.

Jaron Lanier's official site — resources related to the book: https://www.jaronlanier.com/tenarguments.html
Penguin Random House: https://www.penguinrandomhouse.com/books/566677/ten-arguments-for-deleting-your-social-media-accounts-right-now-by-jaron-lanier/
Goodreads: https://www.goodreads.com/book/show/37830765-ten-arguments-for-deleting-your-social-media-accounts-right-now

Eli Pariser — The Filter Bubble: What the Internet Is Hiding from You (2011) Introduced the concept of the filter bubble into the global vocabulary — personalisation algorithms do not show a more relevant world but a smaller one, from which uncomfortable perspectives have disappeared.

Penguin Random House: https://www.penguinrandomhouse.com/books/309214/the-filter-bubble-by-eli-pariser/
Wikipedia Eli Pariser (EN): https://en.wikipedia.org/wiki/Eli_Pariser
Google Books (preview): https://books.google.com/books/about/The_Filter_Bubble.html?id=Qn2ZnjzCE3gC

Evgeny Morozov — To Save Everything, Click Here: The Folly of Technological Solutionism (2013) A critique of the ideology that legitimises tech corporations' behaviour: solutionism — the belief that every human problem has a digital solution, that efficiency is a moral end in itself.

LSE Research Online (academic review): https://eprints.lse.ac.uk/53978/
PublicAffairs Books: https://www.publicaffairsbooks.com/titles/evgeny-morozov/to-save-everything-click-here/9781610391764/

Cal Newport — Digital Minimalism (2019) Documents the phenomenon of technological lock-in and argues for the deliberate, selective use of digital tools.

Author's page: https://www.calnewport.com/books/digital-minimalism/

Hannah Arendt — The Origins of Totalitarianism (1951) A foundational analysis of the concentration of power and the elimination of the capacity for independent thought — the basis of any authoritarian system.

Wikipedia (EN): https://en.wikipedia.org/wiki/The_Origins_of_Totalitarianism
Internet Archive (digital edition): https://archive.org/details/originsoftotalit00aren

Jean Baudrillard — Simulacra and Simulation (1981) Conceptualised 'simulacral imperialism' — the moment in which the representation of a reality becomes more 'real' than the lived reality itself.