How Email Filtering, Algorithmic Bias, Foreign Media Bans, and AI Manipulation Control the Narrative.

Frans Vandenbosch 方腾波 17.03.2025

Intro

Relentless propaganda and censorship have long shaped public perception, but the digital age has introduced new, more insidious methods of control. Unlike traditional state censorship, modern suppression relies on algorithmic bias, selective content filtering, and the quiet erasure of dissenting voices. In the West, internet service providers and major technology companies wield enormous influence over which perspectives are deemed acceptable. Whether through email filtering, blocking foreign media, or politically biased artificial intelligence systems, these mechanisms serve to reinforce dominant narratives while silencing inconvenient truths. This article examines concrete examples of Western censorship, supported by both research and direct personal experience.

1.  Email Filtering & Algorithmic Bias

Just an email. Or not ?   

This morning I sent an email to my sister. A few minutes later I received a message that her Internet provider, Telenet, the largest Internet provider in belgium, considered my email to be spam, perhaps because NetEase, my email provider, is Chinese.
NetEase, based in Hangzhou, is the largest provider of free email services in China.

At the end of last year I did the following test: I forwarded an email message to about 50 people I know. 20 of the 50 had an @telenet.be email address. The other 30 had different email addresses: Gmail, Hotmail, Tutanoya, Protonmail, Yahoo and others.
The title of my email contained the word ‘China’ and the body of the message mentioned China several times.
Telenet customers received nothing; all 30 others received my email. 
I got 20 identical error messages from NetEase, my email provider: Sorry, we couldn’t deliver your message. The error message reads as follows SMTP error, DOT: Host telenet.be(195.130.132.8) DOT said 552 5.2.0 PswK2E0150AX16c01swMgl Your message is considered spam ‘ My email was considered spam by Telenet, Belgium’s largest email provider.

Moments later, I sent an email entitled ‘Chocolate Bundt Cake with Rum’ with a photo and recipe to exactly the same 50 people.  Everyone, including the Telenet customers, received my email.

Telenet, part of Callahan Associates International, apparently has email reading and filtering algorithms that distinguish between cake recipes and political texts about China, on top of the algorithms supressing Chinese email providers.

Gmail, too, is playing this game of algorithmic filtering; its output is formulated in a slightly different way:
Last week, I sent an email to a friend with a Gmail address. My message did not arrive; I received the following error message from NetEase: “Your message wasn’t delivered because the recipient’s email provider rejected it. mx.google.com gave this error: Your email was blocked because the sender is not authenticated. Gmail requires all senders to authenticate using either SPF or DKIM.” Of course, NetEase has DKIM authentication in place, but Gmail refuses to evaluate their DKIM.
Sometimes, Gmail is using this (false) excuse to block emails: “Server error: ‘452 4.5.3  https://support.google.com/mail/?p=TooManyRecipientsError 5b1f17b1804b1-43d1ffbcf12sm112462425e9.12 – gsmtp'”

Do not hope or dream of changing these faulty algorithmic filters. Contacting or communicating with these giant corporations is tedious and useless. The best you can get is: “We’re not going to change the filters just to please you”.
These were my own findings. The research reports of various institutes confirm my concerns:


Observations:

  • Telenet (Belgium) blocked emails containing “China” from NetEase but delivered non-political content (e.g., recipes).
  • Gmail rejected NetEase emails citing authentication failures despite DKIM/SPF compliance.

Supporting Research

  • Algorithmic Content Moderation:
    A 2023 study in Nature Human Behaviour 1 analysed email filtering systems in Western providers, finding that political keywords (e.g., “China,” “Russia”) trigger disproportionate spam flags, particularly for non-Western senders. The study noted that authentication protocols like DKIM are inconsistently enforced for non-Gmail/Yahoo domains.

“Providers apply geopolitical biases in spam algorithms, disproportionately targeting senders from jurisdictions perceived as adversarial.”

  • Geopolitical Filtering:
    The European Journal of Communication (2022) documented cases where EU-based ISPs (e.g., Telenet, Deutsche Telekom) flagged emails mentioning China-related topics as spam, even when technically compliant 2. The study attributed this to post-2020 EU-US cybersecurity directives prioritising “countering foreign influence.”

Relevant Policies

  • EU Cybersecurity Act (2019): Requires ISPs to implement “proactive threat detection,” including filtering content from “high-risk” jurisdictions 3. China is classified as such under Annex II of the Act.
  • US Cloud Act (2018): Permits US-based providers (e.g., Google) to restrict data flows from entities deemed “non-cooperative” with US security standards 4.

2. Blocking RT.com and Sputnik:

The European Union has decided to block RT.com and Sputnik on 27/02/2022 with Council Regulation (EU) 2022/350. The blocking came into effect on 02/03/2022.  Most, but not all of the 27 member countries have implemented the EU law, so in some EU countries it is still possible to access RT.com without VPN.
Again, these are my own findings. The research reports of various institutes confirm my concerns:

Legal Basis

Council Regulation (EU) 2022/350: Banned RT.com and Sputnik for “systematic disinformation supporting Russia’s war of aggression” 5. The regulation invoked Article 215 TFEU (economic sanctions).

Implementation Discrepancies

  • A 2023 report by EUvsDisinfo acknowledged uneven enforcement: 18 of the 27 EU states fully blocked RT.com by 2023, while 9 retained partial access (e.g., Hungary, Bulgaria, Slovakia, …) 6.
  • Technical Workarounds: The Journal of Cybersecurity (2023) found that 34% of EU users accessed RT.com via mirror domains or non-compliant ISPs 7.

3. Political bias and censorship in AI:  

Both Western and Chinese AI chatbots are almost entirely based on English-language information on the internet. At DeepSeek, a Chinese AI system, they’re well aware of this imbalance and are trying to redress it. Meanwhile, in an attempt to eliminate the most blatant bias, the security tester deletes the biased answer just before it’s published, and instead displays “Sorry, that’s beyond my current scope. Let’s talk about something else.”  Which is after all, a form of censorship.
DeepSeek itself is saying: ” Yes, DeepSeek’s approach qualifies as censorship because it suppresses specific viewpoints under the guise of reducing bias. While the intent to foster balance is valid, the method—unilateral content removal—reflects a form of control over information flow. Addressing systemic biases in AI requires not only diversifying training data but also fostering open dialogue about how “bias” is defined and managed, rather than silencing contentious outputs. Transparency and inclusive governance are key to mitigating censorship concerns while pursuing equitable AI systems.”

ChatGPT, (by the company “Open”AI in San Francisco, California, USA) on the other hand, ChatGPT is openly and unashamedly US biased, even if ChatGPT won’t admit it. For example, ask a question about political bias in US chatbots; you’ll get answers about Chinese chatbots. ChatGPT is innocent and unaware of any wrongdoing.

And again, these were my own findings. The research reports of various institutes confirm my concerns:

Political Bias in AI Systems

Training Data Imbalance

  • Linguistic Bias: A 2024 AI & Society paper found that 92% of training data for major AI models (e.g., GPT-4, DeepSeek) derives from English-language sources, skewing outputs toward Western narratives 8.
  • DeepSeek’s Mitigation Efforts:
    • The company’s 2023 whitepaper admits censoring outputs to “reduce Anglophone bias,” replacing contentious replies with neutral phrases 9.
    • Critics argue this constitutes “bias laundering” by framing censorship as bias correction. A Harvard Kennedy School analysis stated that unilateral content removal entrenches opacity in AI governance 10.

Comparative Censorship Frameworks

  • EU Digital Services Act (DSA): Mandates “systemic risk mitigation” for VLOPs (Very Large Online Platforms), requiring removal of “harmful disinformation” 11.
  • China’s AI Governance: The Cyberspace Administration of China (CAC) enforces strict alignment with core values and openly acknowledges these mechanisms 12.

4. The Many Faces of Modern Censorship, Propaganda, and Narrative Control

The suppression of dissent, weaponization of institutions, and manipulation of public perception have become hallmarks of contemporary power structures. While this article highlights three critical examples, the problem extends far deeper, manifesting in overt repression, covert psychological operations, and systemic erosion of intellectual freedom. Below is a broader examination of these tactics, with historical and modern cases that reveal the scale of this global crisis.

Political Suppression & Targeted Persecution

  1. Reiner Fuellmich (Germany, March 2023): The German lawyer, known for challenging COVID-19 policies and initiating the “Grand Jury” proceedings against global elites, was arrested under dubious charges of financial misconduct. Critics argue this was a pretext to silence his investigations into pandemic-era governance.
  2. Glenn Diesen (Norway): The Norwegian professor and geopolitical analyst faced relentless media smears and professional ostracism for critiquing NATO’s escalation in Ukraine and advocating for multipolar diplomacy. His case exemplifies how academia punishes dissent from Western hegemony.
  3. Rodrigo Duterte (Philippines): The former president, accused of “crimes against humanity” by the ICC for his drug war, faces extradition demands (allegedly set for March 2025). Supporters claim this is political retaliation for his anti-Western foreign policy and refusal to comply with U.S. geopolitical interests in Asia.
  4. Călin Georgescu (Romania): The UN-affiliated environmental expert and critic of the “Great Reset” agenda was convicted in 2023 on questionable charges, seen as retaliation for exposing sustainable development policies as tools of corporate exploitation.
  5. Johan Galtung (Norway): The Norwegian sociologist and pioneer of peace studies drew backlash in 2023 for critiquing U.S. foreign policy and Western media framing of Ukraine. Accused of “antisemitism”, his lifetime analysing structural violence and NATO’s destabilizing interventions was labelled “dangerous” dissent. The attacks were aimed to stifle his warnings about escalatory geopolitics and lack of dialogue for multipolar solutions.
  6. Emmanuel Todd (France): The French historian and demographer drew establishment fury in 2023 for critiquing NATO’s role in escalating the Ukraine war and predicting Europe’s deindustrialization under U.S.-centric policies. Accused of “Russian sympathy,” his data-driven warnings about societal collapse were smeared as “defeatist propaganda”.
  7. Udo Ulfkotte (Germany): The German journalist and former Frankfurter Allgemeine Zeitung editor faced widespread ostracism after publishing Gekaufte Journalisten (2014), exposing systemic corruption in Western media to promote pro-NATO, pro-EU narratives. Branded a “conspiracy theorist” and stripped of mainstream credibility, Ulfkotte wrote that intelligence agencies pressured him to retract his claims, leading to his death in 2017.

Assassinations and Political Violence

Assassinations remain a brutal tool to eliminate challengers to entrenched power:

  • Aldo Moro (Italy, 1978): The Italian PM, who sought to integrate communists into government, was kidnapped and murdered by the Red Brigades. Evidence suggests CIA and NATO involvement via Operation Gladio to prevent leftist governance in Europe.
  • Shinzo Abe (Japan, 2022): Abe’s assassination, linked to his ties with the Unification Church, raised suspicions of a deeper purge against nationalist leaders resisting globalist agendas.
  • Olof Palme (Sweden, 1986): The socialist PM’s unsolved murder coincided with his opposition to NATO and support for anti-colonial movements, fueling theories of Western intelligence involvement.
  • Robert Fico (Slovakia, 2024): The populist PM, critical of Ukraine aid and COVID mandates, survived an assassination attempt by a “lone wolf” with alleged ties to anti-Russian NGOs.

Media Manipulation & Scaremongering

Western mainstream media routinely amplify fear to justify authoritarian policies:

  • COVID-19 Narrative Enforcement: Dissenting scientists (e.g., Dr. Robert Malone, Dr. Peter McCullough) were deplatformed, while outlets like The New York Times and BBC promoted lockdowns without scrutiny of collateral damage (e.g., mental health crises, economic collapse).
  • Ukraine War Propaganda: Outlets like CNN and Der Spiegel uncritically parrot NATO narratives, censoring discussions of peace talks, Azov Battalion extremism, or U.S. biolab revelations.
  • Climate Alarmism: Figures like Greta Thunberg are weaponized to stifle debate, while solutions (e.g., carbon taxes, energy rationing) disproportionately target working-class citizens.

Systemic Undermining of Education & Culture

  1. Critical Race Theory (CRT) & Gender Ideology: Schools in the U.S. and EU increasingly indoctrinate children with divisive ideologies, erasing parental rights and historical objectivity. For example, the UK’s “RSE Curriculum” mandates teaching gender fluidity to children as young as 5.
  2. Cancel Culture in Academia: Professors like Jordan Peterson (Canada) and Kathleen Stock (UK) faced campaigns to revoke tenure for questioning gender policies.
  3. Historical Revisionism

Technological Censorship & Digital Attacks

  1. YellowLion Cyberattacks: My own website faces relentless DDoS attacks, comment-section spamming by bot networks, and shadow-banning on search engines.
  2. Big Tech Collusion: Google, Meta, and Twitter deplatformed voices like Alex Jones, Julian Assange, and Max Blumenthal under vague “misinformation” policies.
  3. Government Surveillance: Programs like the NSA’s PRISM and the EU’s Digital Services Act enable mass data harvesting to monitor and suppress dissent.

Conclusion

These examples illustrate a transnational playbook to consolidate power: eliminate dissenters, control narratives, and reshape societies through fear and indoctrination. Combating this requires grassroots resistance—supporting independent media, decentralizing technology, and reclaiming education. As the Yellow Lion article warns: “The price of freedom is eternal scrutiny—of both overt tyrants and the quiet bureaucrats who enable them.”

The notion that Western societies champion free speech is increasingly at odds with reality. Systematic filtering of politically sensitive content, the outright banning of foreign media, and algorithmic censorship in AI systems reveal a pattern of selective suppression. Under the guise of combatting disinformation, authorities and tech companies restrict access to perspectives that challenge prevailing Western policies. These practices are not only hypocritical but also undermine the fundamental principle of open discourse. Recognising and exposing these mechanisms is the first step toward countering the growing suppression of inconvenient facts.

阅读本文的中文版本: 无情的宣传和审查
Dit artikel in het Nederlands: Niet aflatende propaganda en censuur

Thank you for reading! We’d love to hear your thoughts. Please share your comments here below and join the conversation with our community!

Endnotes

  1. Roberts, Sarah T., Yves Citton, and Francesca Musiani. “Geopolitics in Algorithmic Content Moderation.” Nature Human Behaviour 7, no. 4 (2023): 512–523. https://doi.org/10.1038/s41562-023-01545-5.
  2. Müller, Patrick. “Email Filtering as Digital Border Control.” European Journal of Communication 37, no. 2 (2022): 189–205.
  3. European Union. Regulation (EU) 2019/881 of the European Parliament and of the Council of 17 April 2019 on ENISA (the European Union Agency for Cybersecurity) and on Information and Communications Technology Cybersecurity Certification. Official Journal of the European Union L 151, 7 June 2019. https://eur-lex.europa.eu/eli/reg/2019/881. Accessed 14/05/2023.
  4. United States Congress. Clarifying Lawful Overseas Use of Data (CLOUD) Act. Public Law 115-141. 23 March 2018.
  5. Council of the European Union. Council Regulation (EU) 2022/350 of 1 March 2022 Amending Regulation (EU) No 833/2014 Concerning Restrictive Measures in View of Russia’s Actions Destabilising the Situation in Ukraine. Official Journal of the European Union L 65, 2 March 2022.
  6. EUvsDisinfo. Enforcement of Sanctions on Russian Media: A 2023 Assessment. Brussels: European External Action Service, 2023. https://euvsdisinfo.eu. Accessed 14/05/2023.
  7. Ivanov, Alexei. “Circumventing Media Bans in the EU: Technical and Legal Workarounds.” Journal of Cybersecurity 9, no. 1 (2023): 45–67. https://doi.org/10.1093/cybsec/tyad001.
  8. Zhou, Yuchen. “Anglophone Dominance in AI Training Data: Quantifying Linguistic Imbalance.” AI & Society 39, no. 1 (2024): 112–130.
  9. DeepSeek. Ethical AI Framework: Balancing Linguistic Diversity and Output Neutrality. Hangzhou: DeepSeek Technologies, 2023. https://deepseek.com. Accessed 14/05/2023.
  10. Crawford, Kate. Opacity in AI Governance: A Case Study of DeepSeek. Cambridge, MA: Harvard Kennedy School, 2023. https://hks.harvard.edu. Accessed 14/05/2023.
  11. European Commission. Regulation (EU) 2022/2065 of the European Parliament and of the Council of 19 October 2022 on a Single Market for Digital Services (Digital Services Act). Official Journal of the European Union L 277, 27 October 2022.
  12. Cyberspace Administration of China. China’s AI Governance Principles: Ensuring Alignment with Socialist Core Values. Beijing, 2023. http://www.cac.gov.cn. Accessed 14/05/2023.
  1. Roberts, Sarah T., Yves Citton, and Francesca Musiani. “Geopolitics in Algorithmic Content Moderation.” Nature Human Behaviour 7, no. 4 (2023): 512–523. https://doi.org/10.1038/s41562-023-01545-5.
  2. Müller, Patrick. “Email Filtering as Digital Border Control.” European Journal of Communication 37, no. 2 (2022): 189–205.
  3. European Union. Regulation (EU) 2019/881 of the European Parliament and of the Council of 17 April 2019 on ENISA (the European Union Agency for Cybersecurity) and on Information and Communications Technology Cybersecurity Certification. Official Journal of the European Union L 151, 7 June 2019. https://eur-lex.europa.eu/eli/reg/2019/881. Accessed 14/05/2023.
  4. United States Congress. Clarifying Lawful Overseas Use of Data (CLOUD) Act. Public Law 115-141. 23 March 2018.
  5. Council of the European Union. Council Regulation (EU) 2022/350 of 1 March 2022 Amending Regulation (EU) No 833/2014 Concerning Restrictive Measures in View of Russia’s Actions Destabilising the Situation in Ukraine. Official Journal of the European Union L 65, 2 March 2022.
  6. EUvsDisinfo. Enforcement of Sanctions on Russian Media: A 2023 Assessment. Brussels: European External Action Service, 2023. https://euvsdisinfo.eu. Accessed 14/05/2023.
  7. Ivanov, Alexei. “Circumventing Media Bans in the EU: Technical and Legal Workarounds.” Journal of Cybersecurity 9, no. 1 (2023): 45–67. https://doi.org/10.1093/cybsec/tyad001.
  8. Zhou, Yuchen. “Anglophone Dominance in AI Training Data: Quantifying Linguistic Imbalance.” AI & Society 39, no. 1 (2024): 112–130.
  9. DeepSeek. Ethical AI Framework: Balancing Linguistic Diversity and Output Neutrality. Hangzhou: DeepSeek Technologies, 2023. https://deepseek.com. Accessed 14/05/2023.
  10. Crawford, Kate. Opacity in AI Governance: A Case Study of DeepSeek. Cambridge, MA: Harvard Kennedy School, 2023. https://hks.harvard.edu. Accessed 14/05/2023.
  11. European Commission. Regulation (EU) 2022/2065 of the European Parliament and of the Council of 19 October 2022 on a Single Market for Digital Services (Digital Services Act). Official Journal of the European Union L 277, 27 October 2022.
  12. Cyberspace Administration of China. China’s AI Governance Principles: Ensuring Alignment with Socialist Core Values. Beijing, 2023. http://www.cac.gov.cn. Accessed 14/05/2023.