The following dialogue shows how control of the people at scale is still done roughly the same way. But has become decentralised in such a way that its hard to detect and feels like entertainment.
QUESTION: Tell me about the following: TikTok propaganda, neural data harvesting, AI-generated leaders—showing how the mechanisms have mutated, not vanished?
The old tools of influence and control did not disappear. They adapted. Classic propaganda (state newspapers, radio broadcasts, posters, and human agents) relied on mass messaging and relatively crude data collection. Today those functions have mutated into platform algorithms, intimate neural signals, and synthetic personas—more scalable, personalized, harder to attribute, and embedded in everyday consumer technology.
TikTok as algorithmic propaganda infrastructure
TikTok (and its Chinese counterpart Douyin) illustrates the shift from top-down state media to “decentralized” and algorithmically optimized influence. Multiple studies (including peer-reviewed work from the Network Contagion Research Institute and academic analyses) show that searches for sensitive topics such as Tiananmen, Uyghurs, Tibet, or Xinjiang return far less critical content on TikTok than on Instagram or YouTube. Pro-China or CCP-favorable material is amplified even when user engagement favors the opposite. Heavier TikTok use correlates with more positive views of China’s human rights record and desirability as a destination.
The platform also functions through influencers. Authoritarian regimes outsource messaging to foreign creators who build organic-looking audiences; experimental evidence indicates these influencers shift favorability toward China more effectively than official state media, which often triggers backlash. Content about trade, tariffs, “shopping in China,” or everyday life on the mainland frequently aligns with official narratives while feeling like lifestyle or commerce content. Even after U.S. legislative pressure and reported ownership arrangements, concerns persist about algorithm control and residual Chinese influence.
Mutation: Propaganda is no longer a discrete broadcast. It is woven into the recommendation engine, engagement metrics, and short-form culture so that users experience it as entertainment or peer content rather than official messaging.
Neural data harvesting: surveillance moving inside the skull
Brain-computer interfaces (BCIs) and consumer neurotech (EEG headbands for meditation, focus, gaming, etc.) collect raw neural signals. These signals can reveal emotional states, cognitive patterns, mental-health indicators, and, with sufficient data and AI decoding, aspects of intention or recognition that go beyond what behavioral data (clicks, location, purchases) previously provided. Senators and researchers have warned that many consumer neurotech companies retain broad access to this data, share it with third parties under vague policies, and offer limited user rights to revoke or delete it.
Unlike conventional biometrics, neural data is dynamic and highly intimate; it can be used for profiling, targeted influence, or inference even when “anonymized.” Regulatory frameworks lag: some U.S. states have begun classifying neural data as sensitive personal information, but broader protections for cognitive liberty remain incomplete. Workplace or consumer deployments raise additional risks of continuous monitoring of attention or emotion.
Mutation: Classic surveillance harvested external behavior and communications. Neural harvesting reaches toward internal mental states, turning the brain itself into a data source. The mechanism of extraction has moved from observation of actions to potential decoding of the substrate of thought and feeling.
AI-generated leaders and synthetic political personas
Generative AI has produced deepfakes, “glowfakes” (positive fabrications), fanfakes, foefakes, and fully synthetic influencers or avatars that speak, campaign, and mobilize. These range from realistic clones of real politicians delivering speeches they never gave, to entirely fabricated personas (e.g., AI military figures, AI political influencers promoting parties, or synthetic “ordinary people” endorsing policies). Campaigns have used them for both promotional and attack purposes across elections in multiple countries. Some are obvious and meme-like (“slopaganda”); others aim at higher realism. Coordinated networks have deployed AI-generated military or political content at scale on platforms including TikTok.
AI also enables precision tailoring, rapid multilingual variation, and continuous operation without human fatigue or accountability. Synthetic leaders can test messages, flood hashtag spaces, and create the appearance of grassroots support or elite consensus.
Mutation: Traditional influence used human spokespeople, forged documents, or staged events. Now the “leader” or “citizen voice” itself can be generated, scaled, and iterated algorithmically. Authenticity becomes harder to verify, and the volume of synthetic content can drown out organic discourse.
The through-line
These three developments are not isolated gadgets. They represent continuous evolution of the same underlying mechanisms:
- Persuasion moved from broadcast + repetition to algorithmic amplification + influencer laundering + synthetic media.
- Surveillance moved from external records to behavioral traces to neural signals.
- Authority and voice moved from human agents to platform-mediated creators to AI-generated personas.
The result is influence that feels more ambient, personal, and difficult to isolate as “propaganda.” Detection, attribution, and regulation lag behind the technical and commercial incentives. The mechanisms have not vanished; they have become more efficient, more intimate, and more tightly integrated into the digital environment most people inhabit daily.
QUESTION: But the tiktok algorithm seems to have shifted the controlling power to the private sector?
Not quite — the shift is more blurred and hybrid than a clean handoff to the private sector.
ByteDance is "private" but not independent
ByteDance (TikTok’s parent) is a private company, but it operates under significant Chinese Communist Party (CCP) structural influence. Reports and lawsuits from former executives describe Party committees inside the company with access to data and input on operations. Chinese law requires tech firms to support national security and propaganda goals, and ByteDance has faced direct guidance on content. The algorithm isn’t purely neutral commercial software — it exists in an environment where the state can demand cooperation, data access, or adjustments.
This is classic state-capital fusion: the company pursues profit and engagement (private incentive), but the regulatory and political environment in China aligns those incentives with state objectives on sensitive topics.
Algorithmic control in practice
- Commercial logic + state alignment: The algorithm maximizes watch time and engagement. Pro-CCP or China-positive content that performs well gets boosted. Studies repeatedly show systematic differences in content visibility (less critical material on sensitive issues) compared to other platforms. This can emerge from training data, content moderation policies, or explicit guidance rather than crude manual intervention.
- Even after U.S. deals: Reported arrangements (joint ventures, licensing) have raised ongoing concerns that ByteDance could retain influence over the core recommendation algorithm. Critics argue this keeps a backdoor for shaping what Americans see, even under new ownership structures.
Broader picture: Private platforms as vectors
Many authoritarian (and democratic) governments have realized it’s more effective to work through private platforms than replace them:
- Platforms self-censor or adjust algorithms for market access (China is a huge market).
- Governments pressure, regulate, or partner with companies (data requests, content removal orders, “voluntary” guidelines).
- Commercial incentives (engagement farming) often amplify divisive or state-friendly content naturally, reducing the need for heavy-handed orders.
In short, the private sector holds the technical levers (recommendation systems, data, moderation), but state power shapes the environment in which they operate — especially for companies with roots or major interests in authoritarian jurisdictions. The control didn’t fully transfer; it became distributed, harder to see, and often mutually reinforcing between company metrics and government goals.
This is why many analysts describe modern digital influence as hybrid warfare or platform-state entanglement rather than a simple privatization of propaganda. The algorithm gives the private company day-to-day power, but the underlying incentives and oversight mean governments (particularly the CCP in TikTok’s case) retain significant indirect leverage.
QUESTION: It's true about private versus public sector propaganda control. Even Meta and google helped the state to disinform during the pandemic and Jan 6th. So does this signal that there is no fundamental difference between public and private sector. One is just an arm of the other.
There are real differences, but the entanglement is undeniable and growing.
Fundamental differences still exist
- Incentives: Private companies (Meta, Google, ByteDance, etc.) are driven primarily by profit, growth, and shareholder value. Engagement, ad revenue, and user retention rule. They respond to markets, lawsuits, advertisers, and public backlash. Governments pursue power, stability, ideology, and re-election (or regime survival). They can tax, regulate, jail, or shut down platforms.
- Accountability and exit: You can delete apps, switch platforms, or (in theory) influence companies via boycotts and competition. Governments have a monopoly on legitimate coercion. Private firms can be sued or regulated; states write the rules.
- Competition and innovation: Private sectors compete (even oligopolistically). This can produce better tools but also surveillance capitalism. Governments tend toward monopoly or heavy control.
In liberal democracies, these differences create friction — platforms sometimes push back against government demands, leak information, or lobby against regulation. In more authoritarian systems (e.g., China), the distinction collapses more completely.
The entanglement you’re pointing to is real
During the pandemic and around January 6, platforms like Meta, Google/YouTube, and Twitter (pre-Musk) coordinated with government agencies, public health officials, and NGOs on content moderation, labeling, deboosting, and removal. This included flagging "misinformation," suppressing certain stories (e.g., lab-leak discussions early on), and amplifying official narratives.
Similar patterns occur with elections, foreign influence, and national security. Reasons include:
- Regulatory pressure and "voluntary" partnerships: Platforms avoid harsher laws or antitrust action by cooperating.
- Shared elite networks: Revolving doors between tech, government, academia, and NGOs.
- Legal gray zones: Section 230 in the US gives platforms discretion (and protection) to moderate; governments exploit that.
- Risk aversion: Companies fear being blamed for real harms (violence, public health crises) or losing access to key markets.
This creates a public-private partnership model for information control that looks like an "arm" relationship on certain issues. Critics on both left and right have called it out — censorship industrial complex, state-adjacent tech, etc.
But it's not total fusion — and that matters
- Private companies also resist when it hurts their business (e.g., Apple pushing back on some encryption demands, Musk's changes at X opening up discourse, or platforms fighting EU overreach).
- Users and markets can punish overreach (trust erosion, migration to alternatives like Rumble, Substack, or decentralized options).
- Different administrations and countries pull in different directions. What one U.S. administration pressures platforms to suppress, the next might criticize as viewpoint discrimination.
- In China, the fusion is deeper by design (Party cells in companies, real-time data access). In the West, it's messier, contested, and leaky — which is why leaks, lawsuits, and whistleblowers keep exposing it.
Bottom line: There is a meaningful difference between public and private in principle (coercion vs. contract, monopoly vs. competition). In practice, especially around high-stakes information flows, the sectors often align or collude because it serves both their interests. Private platforms amplify state narratives when convenient and suppress when pressured — not always as a formal "arm," but as interdependent players in the same ecosystem.
This hybrid reality is why many see digital propaganda/surveillance as more insidious than old-school state media: it wears the clothes of private entertainment and "community standards" while channeling power. The solution isn't pretending they're identical, but increasing transparency, competition, user choice, and clear limits on government jawboning of platforms.