Chinese Bot Farms and AI Data Centers: The Attack on USA Infrastructure
- X removed approximately 200,000 inauthentic accounts linked to China, 200 of which focused on anti-AI data center campaigns.
- The strategy consisted of amplifying real grievances (electricity costs and grid stress) through fake news and AI-generated content.
- The operation follows a similar precedent by OpenAI in June 2026, confirming a pattern of foreign influence.
- The strategic goal is potentially the slowing down of the United States AI infrastructural build-out.
X's operation: 200,000 accounts and the information war
On August 28, 2026, the X Global Government Affairs team announced the identification and disruption of a vast network of inauthentic accounts, estimated at approximately 200,000 profiles, attributable to Chinese operators. Although the critical mass of the bot farm was imposing, attention focused on a specific core of 200 accounts dedicated to a precise mission: manipulating US public perception regarding artificial intelligence data centers and related energy policies.
This was not a traditional cyberattack aimed at compromising systems or hacking the power grid, but a sophisticated influence campaign. The accounts operated by posing as ordinary American citizens, using names such as Thomas Jackson, Patricia Robinson, Sandra Thompson, and Kenneth Flores to give a human and local face to the propaganda.
Manipulation tactics
The strategy used by the bot farm was not based on total lies, but on the amplification of real grievances to make them more toxic and polarizing. The content focused on three narrative pillars:
- Cost increase: the claim that AI data centers are pushing up household electricity bills.
- Infrastructural stress: the thesis that the speed of construction of these facilities is putting the stability of the US power grid at risk.
- Economic inequality: the narrative that ordinary families are effectively subsidizing extremely wealthy AI and data center companies.
To give credibility to these theses, the bots used material from legitimate reports, such as an article from the Southern Maryland Chronicle regarding the increase in capacity costs of PJM Interconnection. This real material was paired with manipulative content: AI-generated cartoons and comic strips depicting data center developers as villains getting rich while families suffer from energy costs.
Context and Precedents: a pattern of interference
The event is not an isolated case. X and other tech companies report a recurring pattern of foreign influence operations targeting the AI sector. In particular, X's operation closely follows a similar takedown carried out by OpenAI in June 2026. The correlation was confirmed by the fact that at least two of the images used by the Chinese bot farm had previously appeared in OpenAI's report.
The technical-economic context makes these campaigns particularly effective. According to a June 2026 market update from PJM (which serves 13 states and Washington, D.C.), data centers are currently the primary driver of new energy demand growth, as these facilities can be built two to three times faster than other industrial infrastructures.
X described the activity as an attempt to interfere with what the platform itself recognized as a 'legitimate debate' on energy policies and American AI.
Strategic Analysis: Why target Data Centers?
[GLACOM.NEWS ANALYSIS]: From a geopolitical point of view, China's goal is not simply to create social chaos, but to hit the physical bottleneck of AI: energy. Artificial intelligence is not just software, but heavy infrastructure. If the US public perceives data centers as enemies of the domestic wallet, political pressure will lead to:
- Slowing down of construction permits (permitting).
- New restrictive regulations on energy consumption.
- More aggressive local opposition (NIMBY - Not In My Backyard).
By slowing down the US infrastructural build-out, China could gain precious time to accelerate its own race toward innovation, transforming an internal energy management problem into a global strategic competitive advantage.
Impact on the Market and the Tech Sector
The impact of these operations is reflected on several market levels:
1. Reputational Risk for Developers
Companies building data centers are painted as predatory entities. This can influence ESG ratings and complicate relationships with local administrations, making land acquisition and grid connection more expensive and slower.
2. Pressure on Energy Utilities
The amplification of discontent shifts attention to utilities (such as PJM), which find themselves having to manage not only the technical challenge of the energy load, but also increasing political pressure to protect consumer prices at the expense of industrial growth.
3. Security of Social Platforms
The case highlights how platforms like X have become the main battlefield for cognitive warfare. The ability to identify 200,000 accounts suggests an improvement in detection systems, but the persistence of these campaigns indicates high-level state determination.
What it means for an Entrepreneur
[GLACOM.NEWS ANALYSIS]: For the entrepreneur operating in tech, energy, or AI, this dossier highlights three critical risks to include in their corporate risk matrix:
- Risk of "Grievance Harmonization": Companies must be aware that legitimate criticisms (e.g., environmental or energy impact) can be "hijacked" and amplified by foreign actors for political purposes. Corporate communication can no longer be just technical, but must be proactive and transparent to prevent the information vacuum from being filled by bot farms.
- Infrastructural Dependence: The vulnerability of the power grid is not just a technical problem, but a geopolitical risk. Investments in proprietary energy or off-grid renewables for data centers are no longer just a sustainability choice, but a political risk mitigation strategy.
- Vigilance on Data and Opinion: Those managing AI products must monitor how their sector is perceived online. The manipulation of public perception can quickly translate into legislative decisions that impact the business model (e.g., taxes on AI energy consumption).
Future Scenarios: Toward an "Energy Cold War"
Based on the reported facts, we can hypothesize three scenarios for the coming months:
Scenario A: Escalation of Propaganda
China could diversify targets, shifting attention from electricity costs to themes such as the water impact of data centers or job losses, using generative AI to create even more convincing and difficult-to-trace content.
Scenario B: USA Legislative Reaction
The US government could respond with more aggressive incentives for energy dedicated to AI, seeking to separate the industrial grid from the domestic one to neutralize the "costs for families" argument used by bot farms.
Scenario C: Strengthening of Tech Defense
Closer collaborations between X, OpenAI, and intelligence agencies to create an early warning system against influence campaigns, transforming content moderation into a national security function.
In conclusion, X's operation reveals that the race for AI is not fought only in research labs or chip factories, but also in the public perception of who pays the electricity bill. The ability to build infrastructure will increasingly depend on the ability to win the battle of the narrative.
FAQ
How many accounts were removed by X?
X disrupted approximately 200,000 inauthentic accounts linked to China, 200 of which were specifically dedicated to anti-AI data center campaigns.
What was the main objective of the Chinese bot farm?
To manipulate the American public debate on AI and energy, claiming that AI data centers increase domestic electricity bills and overload the grid.
What tools did the bots use to influence users?
They used fictitious American citizen names, citations from real reports (such as those from PJM Interconnection), and AI-generated images/comics depicting data center operators as greedy.
Have there been similar episodes in the past?
Yes, OpenAI carried out a similar takedown in June 2026, and some images used by the X bot farm had already appeared in OpenAI's report.
Sources: Startupfortune, Nextbigfuture, Techspot, Yahoo, Tech · dossier by glacom's AI newsroom
.png)
