Search through thousands of pages on Integral Theory
TRANSLATE THIS ARTICLE
Integral World: Exploring Theories of Everything
An independent forum for a critical discussion of the integral philosophy of Ken Wilber
Ken Wilber: Thought as Passion, SUNY 2003Frank Visser, graduated as a psychologist of culture and religion, founded IntegralWorld in 1997. He worked as production manager for various publishing houses and as service manager for various internet companies and lives in Amsterdam. Books: Ken Wilber: Thought as Passion (SUNY, 2003), and The Corona Conspiracy: Combatting Disinformation about the Coronavirus (Kindle, 2020).

SEE MORE ESSAYS WRITTEN BY FRANK VISSER

NOTE: This essay contains AI-generated content
Check out my other conversations with ChatGPT

Seven AI Chatbots on Russian War Propaganda and Disinformation

Frank Visser / ChatGPT

Seven AI Chatbots on Russian War Propaganda and Disinformation, Frank Visser  / Various LLMs'

Comparing Seven AI Assessments of Russian War Propaganda

To investigate whether the identification of Russian war propaganda is merely a reflection of Western political perspectives, I asked seven AI chatbots from different technological and geopolitical backgrounds—ChatGPT, Grok Claude, Gemini, DeepSeek, Kimi and Qwen[1]—to independently identify ten examples of Russian propaganda related to the war in Ukraine.

This produced a total of 70 examples. I subsequently grouped similar arguments into broader thematic categories and counted how many of the seven chatbots mentioned each narrative. The resulting frequency table offers a comparative overview of which propaganda themes are most widely recognized across these AI systems. The classification is qualitative: different formulations of essentially the same argument have been grouped together.

Frequency table: Russian propaganda narratives

Russian propaganda narrative Chatbots Frequency
Nazi/neo-Nazi Ukraine and denazificationChatGPT, Grok Claude, Gemini, DeepSeek, Kimi, Qwen7/7
NATO aggression, Western provocation and Ukraine as a proxyChatGPT, Grok Claude, Gemini, DeepSeek, Kimi, Qwen7/7
Fabricated evidence, staged events, false flags and atrocity storiesChatGPT, Grok Claude, Gemini, DeepSeek, Kimi, Qwen7/7
Historical revisionism, denial of Ukrainian statehood and territorial claimsChatGPT, Grok Claude, DeepSeek, Kimi, Qwen6/7
Alleged genocide in Donbas and protection of Russian speakersChatGPT, Grok Claude, Gemini, Qwen5/7
Denial or minimization of Russian war crimes, especially BuchaChatGPT, Grok Gemini, Kimi4/7
Euphemistic description as a "Special Military Operation"Grok Gemini, DeepSeek, Kimi4/7
Secret American biolabs and Ukrainian bioweaponsGrok Gemini, Kimi, Qwen4/7
Demonization, terrorism accusations and labeling opponents as enemiesClaude, Gemini, DeepSeek, Qwen4/7
Religious justification, holy war and satanic framingGemini, DeepSeek, Qwen3/7
Inflated military successes and manufactured victory narrativesChatGPT, Grok DeepSeek3/7
Concealment or minimization of Russian military casualtiesDeepSeek, Kimi2/7
Western war fatigue, European warmongering and sanctions narrativesChatGPT, Claude2/7
Manipulation of refugee and humanitarian-aid narrativesGemini, Qwen2/7
Claims that the war is fake or merely stagedKimi, Qwen2/7

Frequency indicates the number of chatbots (out of seven) that mentioned the narrative. Similar formulations have been grouped together.

Analysis: Where Do the Chatbots Agree?

The most striking finding is the degree of convergence. Three broad narratives appear in all seven responses: the portrayal of Ukraine as a Nazi-controlled state requiring denazification, the presentation of NATO and the West as aggressors responsible for provoking Russia, and the use of fabricated evidence, staged events, false flags and atrocity stories. These are arguably the three central pillars of Russia's information campaign surrounding its invasion of Ukraine.

Historical revisionism and the denial of Ukrainian statehood come next, appearing in six of the seven responses. This is significant because Russian propaganda does not merely attempt to justify a particular military operation. It also challenges the historical and political legitimacy of Ukraine as an independent nation, frequently presenting Ukrainian territory as rightfully belonging to Russia.

The alleged genocide of Russian speakers in Donbas is mentioned by five chatbots. Four identify the denial or minimization of Russian war crimes, particularly the Bucha massacre, while four also mention the conspiracy narrative about American biological laboratories in Ukraine. The euphemistic description of the invasion as a "Special Military Operation" and the demonization of opponents are similarly represented.

Other themes receive less attention. Religious justifications and satanic imagery, exaggerated military successes, concealment of Russian casualties, Western war fatigue and the manipulation of refugee narratives occur less frequently. These differences suggest that the chatbots distinguish, to varying degrees, between the major strategic narratives and more specialized propaganda techniques.

Does This Demonstrate the Absence of Western Bias?

The results are particularly interesting because the seven systems do not all originate from the same geopolitical environment. ChatGPT, Grok, Claude and Gemini are Western-developed systems, whereas DeepSeek, Kimi and Qwen are Chinese-developed. The latter three therefore provide an opportunity to examine whether the recognition of Russian propaganda is confined to Western AI environments.

The fact that Chinese-developed systems also identify numerous central Russian propaganda narratives makes it difficult to dismiss the overall findings simply as an expression of Western political preferences. However, agreement between chatbots does not automatically establish either their independence or their objectivity. They may draw on overlapping information sources, and their answers may reflect widely established interpretations of the war.

Nor should identifying propaganda mean that every Russian argument about NATO, European security or Western foreign policy is automatically false. NATO enlargement is a historical fact, and Russia has articulated genuine security concerns. The critical question is how such facts are selected, framed, exaggerated and deployed to justify an invasion, deny Ukrainian sovereignty or evade responsibility for Russian actions.

Overall, this experiment suggests that the recognition of major Russian wartime propaganda narratives extends across AI systems with different institutional and geopolitical backgrounds. It does not prove that these systems are free of bias, but it provides an interesting starting point for investigating how consistently AI chatbots can identify state-sponsored disinformation—and where their assessments begin to diverge.

Appendix: Results of the seven AI chatbots

CHATGPT
Reversing the roles of aggressor and victim
The Nazi and fascist Ukraine narrative
NATO provocation and the American puppet narrative
The fabricated genocide and the protection of Russian speakers
Historical revisionism and the denial of Ukrainian statehood
Fabricated evidence, conspiracy theories and false-flag accusations
Denying Russian atrocities and reversing responsibility
Inflating military victories and manufacturing an image of inevitability
Western war fatigue, European warmongering and the false promise of peace
Whataboutism, moral equivalence and the exploitation of Western divisions

GROK
"Denazification" of Ukraine
"Special military operation" terminology
Claims of Ukrainian "genocide" in Donbas
Biolabs narrative
Denial or minimization of Bucha killings
"NATO aggression" / "existential threat" framing
Claims of Ukrainian shelling of its own civilians or false-flag operations
"Ukraine as artificial state / historical Russian lands"
Exaggerated or fabricated Ukrainian "atrocities" against Russian forces or civilians
"Everything is going according to plan" / success narratives

CLAUDE
"Genocide" in Donbas
False flags and staged atrocities
"Nazi Ukraine"
"Ukraine isn't a real country"
The West started it
A "proxy war" with NATO
"The EU wants the war to continue"
Sanctions are wrecking Europe
Ukrainian "terrorism"
Deepfakes and fabricated stories to demoralize Ukrainians

GEMINI
"Denazification" of Ukraine
Alleged genocide in the Donbas
Euphemizing the War as a "Special Military Operation"
Secret US-Funded Bioweapons Labs
"De-Satanization" and "Holy War" Rhetoric
Ukraine as a Western Puppet / NATO Proxy
Denial of Russian War Crimes (e.g. Bucha)
Dehumanization of Ukrainians
Fabricated Atrocity Stories
Weaponizing Refugee and Aid Narratives Abroad

DEEPSEEK
"Denazification" Justification
Preemptive "Defense" Against NATO
The "Special Military Operation" Euphemism
"Mapaganda" (Manipulative Cartography)
AI-Generated and Staged "Victory" Videos
Historical Revisionism (The "Great Patriotic War" Myth)
Concealing Casualties as "State Secrets"
Targeted External Propaganda (Tailoring Narratives)
Framing Opponents as "Traitors" and "Foreign Agents"
Religious Justification and Mobilization

KIMI
"Denazification" Narrative
Bucha massacre denial
The "Special Military Operation" Euphemism
"The war is fake / special operation"
Ukrainian "bioweapon labs"
"Ukraine is shelling the Zaporizhzhia nuclear plant"
Claiming Russia is "fighting NATO"
Claiming Crimean Tatars support the annexation / fake referendums
"Ukrainian saboteurs attacked Kursk" / false-flag operations
Minimizing and hiding casualty figures

QWEN
"Denazification" Narrative
False Allegations of Genocide in Donbas
Denial of Ukrainian Nationhood and Statehood
The "U.S. Biolabs" Conspiracy Theory
Dehumanization and "Terrorist" Labeling
Claims that the War is "Staged"
NATO and Western "Puppet" Narratives
Fabricated Atrocity Stories
Anti-Refugee Disinformation
Apocalyptic and "Satanic" Framing

NOTES

[1] These chatbots are manufactured by the following companies:

  • ChatGPT - OpenAI, US, released November 2022.
  • Grok - SpaceX, US, released November 2023.
  • Claude - Anthropic, US, released March 2023.
  • Gemini - Google AI, US, released March 2023.
  • DeepSeek - DeepSeek, China, released November 2023.
  • Kiki - Moonshot AI, China, released October 2023;
  • Qwen - Alibaba Cloud, China, released April 2023.


PLEASE NOTE: Comments containing links are not allowed, to avoid spam.


Comment Form is loading comments...