In May 2026, it was disclosed that a Super Political Action Committee (Super PAC) backed by US AI companies had raised $140 million to recruit influencer-bloggers on overseas social media to smear China's AI — at an opening price of $5,000 per video. The episode reveals a structural fact that goes well beyond the level of "lobbying by individual firms": the US opinion war against China's AI has entered a phase of institutionalized operation, complete with a clearly defined funding pool, a standardized two-phase playbook, and a vast content distribution network.
A Two-Phase Script: From Self-Mythologizing to Demonizing the Rival
The Super PAC's opinion-shaping plan was not improvised. It carries the hallmarks of a deliberate two-phase strategic design:
Phase One: hype and deify America's AI development. The core task is to bankroll a flood of content stressing the importance of investing in American AI technology, cultivating the narrative that "only America can lead the AI revolution." The rhetorical engine of this phase is fear marketing — "if we don't spend more, America will be overtaken" — pressed into service to push policymakers and the public toward more aggressive AI investment and export controls.
Phase Two: demonize China's AI. Build a systematic negative narrative around China's AI, with scripted talking points spanning multiple dimensions — from "violating data security" to "stealing jobs," from "technology theft" to "state surveillance." Each video is priced at $5,000, with influential influencer-bloggers recruited to publish on overseas social platforms.
The plan essentially has two axes: the first phase hypes and deifies America's AI development, stressing the importance of investing in American AI technology; the second phase demonizes China's AI. And there are templates for the corresponding talking points, covering everything from "violating data security" to "stealing jobs."
The subtlety of the operation lies in the positive-negative synergy between the two phases: the positive narrative supplies legitimacy for investment and controls, while the negative narrative manufactures urgency for suppression and blockade.
Three Maturity Indicators of Institutionalized Operation
2.1 The Funding Pool: $140 Million as Institutional Investment
A $140 million war chest signals that this is no one-off PR stunt but a continuously running institutional operation. Contributions raised and pledged to the Super PAC have reached a scale sufficient to sustain large-scale, sustained content production. At the quoted rate of $5,000 per video, $140 million covers the production of 28,000 videos — an industrial-grade content output capacity.
2.2 The Systematization of Talking Points: From Lobbying to Content Production
It is hardly new for US AI companies to leverage "security threat" narratives in lobbying the government — as early as DeepSeek's breakout, firms such as OpenAI were already petitioning the White House to amplify the "China threat" and recommend "AI technology export controls." But this episode marks a qualitative leap: the reach extends from lobbying the government to penetrating public opinion directly. The line between lobbying and propaganda is being systematically blurred.
2.3 The Socialization of Content Distribution
The strategy of using influencer-bloggers as distribution nodes exploits social media's trust mechanics — audiences tend to trust "a creator's judgment" more than official media or corporate statements. At $5,000 apiece, the cost is remarkably low compared with traditional advertising, yet in terms of narrative penetration it may deliver a considerably higher return.
A Garden in Full Bloom, Grown Out of the Blockade
To understand the timing and motive of this opinion operation, one needs to return to the trajectory of America's AI blockade against China:
First it banned Nvidia, the dominant AI chipmaker, from exporting high-end AI chips to China; then it pushed OpenAI, the leader in large AI models, to stop opening up to China — under layer upon layer of pressure, China's AI companies found another way, competing their way through architectures and engineering until, at last, they reached "a new village beyond the willow shade and the blooming flowers." — Chang'anjie Zhishi
The effects of the blockade policy have run precisely counter to intent:
- The hardware gap accelerated substitution: after DeepSeek released its V4 preview, at least 8 domestic AI chipmakers simultaneously announced compatibility solutions, offering developers and enterprises a diversified menu of domestic compute options
- The ecosystem matured faster: from "a lone brave" to "a hundred flowers blooming" — foundation models have their "Five Tigers," GPUs their "Four Little Dragons," and tens of thousands of universities, research institutes, companies, and individual developers are pushing forward the "domestic compute + domestic large models" ecosystem
- Objective data corroborates: China counts more than 6,000 AI companies; cumulative global downloads of domestically developed open-source large models have surpassed 10 billion; China's share of global AI patents reaches 60%
Against this backdrop, the escalation of the AI opinion war can be read as the second-best option after the blockade policy failed — when technology blockades and export controls could not halt China's AI development, shifting to the opinion arena to conduct a "cognitive blockade" becomes the substitute strategy.
The Self-Limitation of Narrative: When Experience Outperforms the Script
One signal worth noting: despite the generous price on offer, many American bloggers still refused the work. This supplies one piece of evidence for the self-limitation of narrative efficacy: the actual performance and open-source posture of China's AI products are eroding the persuasive foundation of the negative script. The "do more with less" strategy of Chinese AI products such as DeepSeek — avoiding excessive consumption of compute, opening model architectures, and lowering the cost of deploying AI across society — creates user experiences that directly contradict the negative narrative.
How China's AI really is — global users are using it, experiencing it, and judging it for themselves. Those real impressions cannot be distorted by a few pieces of custom-made smear content. — Chang'anjie Zhishi
The episode also corroborates a broader regularity: in technology domains like AI, where experience directly shapes perception, the effect of opinion operations is bounded by product facts. When the product in users' hands diverges sharply from what the narrative conveys, the efficacy boundary of the narrative makes itself felt.
The "Three-Stage Evolution" Framework of the Tech Opinion War Against China
From this episode one can distill a reusable analytical framework — the "three-stage evolution" of America's tech opinion war against China:
| Stage | Primary Mode of Operation | Target Audience |
|---|---|---|
| Stage One: Lobbying-Driven | Companies and industry associations lobby the government directly, amplifying security threats | Policymakers |
| Stage Two: Media Framing | Mainstream media construct systematic narrative frames (e.g., "China Shock 2.0," "industrial distillation") | Opinion elites |
| Stage Three: Content Industrialization | Super PAC + social media influencers, industrialized content production and distribution | Mass public opinion |
The current episode marks the transition from Stage Two to Stage Three — narrative operation descends from the "elite layer" to the "mass layer", and the mode of content production expands from "professional production by media institutions" to "industrialized outsourced production for social media."
Three chains of evidence — 1) the $140 million institutionalized operation of the Super PAC; 2) the systematic output of the two-phase talking-point templates; 3) the Chinese AI ecosystem that the blockade spawned rather than contained — together point to a new phase of institutionalized operation in America's AI opinion war against China. Yet the technological characteristic that "experience determines perception" sets an objective boundary to the narrative efficacy of such operations.