AI-generated content, including deepfakes, is primarily used by state actors and propagandists to manipulate public perception in geopolitical conflicts.

Verdict: likely FALSE (67%)

Web Report: https://solsice.com/public/debates/ai-generated-content-including-deepfakes-is-primarily-used-b-e41a18284146

---

This section provides a brief overview of the key arguments. You do not need to read the full detailed report below.

✅ Key PRO arguments:

1. ■State actors [33] dominate because advanced generative models [16] require vast computational resources and specialized talent; training a single large language model [25] can cost hundreds of millions of dollars, and generating deepfake videos at scale demands data-center-grade GPU clusters [18].
2. ■In the Russia-Ukraine war [30], a deepfake of President Zelenskyy appeared to call on soldiers to surrender, showing state-linked operations run industrialized influence campaigns [21].
3. ■The opposing side's table claiming entertainment/social media accounts for 65% of AI content and state/political actors 5% is unverifiable—no source, methodology, time period, or volume definition—and misses the geopolitical manipulation domain.

❌ Key ANTI arguments:

1. ■AI-generated content [1] is predominantly driven by decentralized actors [8] and commercial interests rather than state entities; the affirmative substitutes existence of state-linked operations for proof of primary use.
2. ■High-profile incidents [19] like the Zelenskyy deepfake [40] are isolated, high-impact anomalies and do not represent the total volume of AI-generated content [1]; the argument lacks a correlation or transfer operator [39] linking intensity to volume.
3. ■The Spamouflage network [32]'s fake broadcast ecosystems [13] are documented state-sponsored activity, but their existence does not establish that such operations are the primary use; no comparative baseline [4] or statistical transfer [36] shows state-led volume exceeds massive decentralized output.

💭 Conclusion: False, the evidence does not support the claim that AI-generated content [1] is primarily used by state actors [33] for geopolitical manipulation. The estimated volume share for state and political actors is only 5%, while entertainment and social media accounts for 65% and commercial and marketing for 25%.

---

🔬 DeepResearch Result: likely FALSE ❌ (67%)

Assertion: AI-generated content [1], including deepfakes [9], is primarily used by state actors [33] and propagandists [28] to manipulate public perception [29] in geopolitical conflicts [17].

Participating models: deepseek-v4-pro 💬, gemma-4-26b-a4b-it 💬👁️

📊 Face-offs: 0 concluded TRUE, 1 concluded FALSE (1 face-offs, 3 models)
📊 Score by side: TRUE=0.00, FALSE=0.67

🏅 Model score changes:
deepseek-v4-flash-0731 💬: +7

✅ PRO Arguments:

1. ■State actors dominate because advanced generative models [16] require vast computational resources and specialized talent; training a single large language model [25] can cost hundreds of millions of dollars, and generating deepfake videos at scale demands data-center-grade GPU clusters [18]. deepseek-v4-pro 💬
2. ■In the Russia-Ukraine war [30], a deepfake of President Zelenskyy appeared to call on soldiers to surrender, showing state-linked operations run industrialized influence campaigns [21]. deepseek-v4-pro 💬
3. ■The opposing side's table claiming entertainment/social media accounts for 65% of AI content and state/political actors 5% is unverifiable—no source, methodology, time period, or volume definition—and misses the geopolitical manipulation domain. deepseek-v4-pro 💬
4. ■Non-state hobbyists produce isolated, often crude content, while state-linked operations run industrialized influence campaigns. deepseek-v4-pro 💬

❌ ANTI Arguments:

1. ■AI-generated content is predominantly driven by decentralized actors [8] and commercial interests rather than state entities; the affirmative substitutes existence of state-linked operations for proof of primary use. gemma-4-26b-a4b-it 💬👁️
2. ■High-profile incidents [19] like the Zelenskyy deepfake [40] are isolated, high-impact anomalies and do not represent the total volume of AI-generated content; the argument lacks a correlation or transfer operator [39] linking intensity to volume. gemma-4-26b-a4b-it 💬👁️
3. ■The Spamouflage network [32]'s fake broadcast ecosystems [13] are documented state-sponsored activity, but their existence does not establish that such operations are the primary use; no comparative baseline [4] or statistical transfer [36] shows state-led volume exceeds massive decentralized output. gemma-4-26b-a4b-it 💬👁️
4. ■A table indicates entertainment and social media sector [12] accounts for 65% of AI-generated content volume [5] while state and political actors account for only 5%. gemma-4-26b-a4b-it 💬👁️

💭 Reasoning: False, the evidence does not support the claim that AI-generated content is primarily used by state actors for geopolitical manipulation. The estimated volume share for state and political actors is only 5%, while entertainment and social media accounts for 65% and commercial and marketing for 25%.

📋 PRO Facts:
• Training a single large language model can cost hundreds of millions of dollars.
• A deepfake of President Zelenskyy appeared in the Russia-Ukraine war.
• State-linked operations run industrialized influence campaigns.

📋 ANTI Facts:
• The entertainment and social media sector reportedly accounts for 65% of AI-generated content volume; state and political actors account for 5% (unverified).

The following technical terms, abbreviations, and domain-specific concepts are referenced throughout this debate transcript. Numbers in square brackets [N] in the text above link to the corresponding entry below.

[1] AI-generated content — Artificial Intelligence-generated content — Content such as text, images, audio, or video produced by artificial intelligence systems; the debate concerns its use in manipulating public perception.

[2] barrier to entry — The high cost and specialized resources required to produce advanced AI content, which the affirmative argues favors state actors.

[3] commercial enterprises — Businesses that produce AI-generated content for profit, cited by the negative as part of the decentralized ecosystem.

[4] comparative baseline — A reference point needed to compare the scale of state versus non-state AI content; the negative argues the affirmative lacks one.

[5] content volume — The amount of AI-generated content produced, used as a metric in the debate; the negative cites sector shares.

[6] coordinated inauthentic accounts — Social media accounts operated in a coordinated manner to deceive users about their origin, used to amplify state-linked content.

[7] cybersecurity firms — Companies that analyze digital threats and trace disinformation infrastructure; the transcript mentions Sensity AI and Mandiant.

[8] decentralized actors — Individuals and groups not under central state control, whom the negative argues produce most AI content.

[9] deepfakes — Synthetic media, usually video, in which a person's appearance or voice is manipulated by AI to create false events.

[10] democratization of generative AI — The process by which AI content creation tools become widely accessible, lowering the barrier to entry for individuals and small businesses.

[11] disinformation networks — Coordinated infrastructures that spread false information, often linked to state actors.

[12] entertainment and social media sector — The industry producing amusement and social media content, cited by the negative as accounting for 65% of AI content volume.

[13] fake broadcast ecosystems — Networks that mimic legitimate broadcasting operations to spread propaganda, such as the Spamouflage network.

[14] forensic analysis — The investigation of digital content and infrastructure to trace its origin and attribution.

[15] generative AI market — The commercial ecosystem of AI content production driven by businesses and individual users.

[16] generative models — AI systems that create new content, such as text, images, or video, based on training data.

[17] geopolitical conflicts — Disputes between nations or blocs over territory, influence, or ideology, such as the Russia-Ukraine war.

[18] GPU clusters — Graphics Processing Unit clusters — Groups of graphics processing units used together to provide the massive computing power required for AI training and deepfake generation.

[19] high-profile incidents — Notable, widely reported events, such as the Zelenskyy deepfake, used as examples in the debate.

[20] individual creators — Private persons who produce AI content, often using democratized tools.

[21] influence campaigns — Organized efforts to shape public opinion or behavior, often using coordinated content and accounts.

[22] information warfare — The use of information and communication technologies to gain strategic advantage over an adversary.

[23] institutional actors — Large organizations such as governments or corporations, as opposed to individual creators.

[24] Kremlin-aligned media — Media outlets aligned with the Russian government, mentioned as disseminating a deepfake of President Zelenskyy.

[25] large language model — LLM — An AI model trained on vast text data to generate human-like language; training can cost hundreds of millions of dollars.

[26] marketing copy — Promotional text used in advertising, cited as an example of non-geopolitical AI content.

[27] non-state actors — Individuals or organizations not affiliated with a government, whom the negative argues produce most AI content.

[28] propagandists — Individuals or groups that spread biased or misleading information to promote a political cause.

[29] public perception — The collective opinion or understanding of the public, which actors seek to manipulate in geopolitical conflicts.

[30] Russia-Ukraine war — The armed conflict between Russia and Ukraine, used as an example of a geopolitical conflict in the debate.

[31] social media amplifiers — Accounts or systems used to boost the reach of content on social media platforms.

[32] Spamouflage network — A network mentioned in the debate as a documented instance of state-sponsored disinformation activity creating fake broadcast ecosystems.

[33] state actors — Governments or government-linked organizations that operate in a coordinated manner, as opposed to private individuals.

[34] state-linked disinformation networks — Networks connected to a government that spread false information, such as the infrastructure traced in the Zelenskyy deepfake case.

[35] state-sponsored deepfakes — Deepfakes produced or funded by governments for political purposes.

[36] statistical transfer — A logical operation that applies statistics from one context to another; the negative argues the affirmative fails to provide it.

[37] synthetic media — Media content generated or manipulated by AI, including deepfakes.

[38] total AI content volume — The overall quantity of AI-generated content across all sectors, used as a metric in the debate.

[39] transfer operator — A mathematical or logical operation needed to generalize from specific cases to a broader population; the negative argues the affirmative fails to provide it.

[40] Zelenskyy deepfake — A specific deepfake video of Ukrainian President Zelenskyy appearing to call on soldiers to surrender, disseminated by Kremlin-aligned media.

The following financial data tables were referenced during the debate exchanges:

Legend: Estimated distribution of AI-generated content volume by sector. Data represents qualitative market analysis of generative AI output trends.