Chinese state-affiliated actors use algorithms to suppress dissenting content on social media
A Canada-based Chinese blogger reportedly documented a coordinated algorithmic suppression attack targeting their X account in March 2026, using blank accounts and coordinated blocking to exploit the platform's recommendation algorithm. Multiple commentators alleged that Chinese state actors have systematically studied social media algorithms to suppress dissenting content for over two years.
What happened
On 7 March 2026, a Canada-based Chinese blogger published an account of algorithmic suppression targeting their X platform account. The account had grown from 2,800 to 7,800 followers between mid-February and early March 2026, with individual posts reaching as many as 870,000 impressions.
The blogger described a three-stage attack: 'embedding' (blank accounts suddenly following their profile), 'dormancy' (those accounts remaining in the following list while being algorithmically associated with low-quality social graphs), and 'detonation' (coordinated unfollowing and blocking when political posts were gaining traction, with 30+ accounts blocking on a single occasion, concentrated during Beijing timezone working hours).
The blogger analysed the mechanism as exploiting X's recommendation algorithm: when large numbers of followers simultaneously unfollow during peak post engagement, the algorithm misinterprets the content as objectionable or false, triggering automated suppression of distribution to non-followers. Simultaneous blocking signals rule violations, placing accounts on search blacklists.
Reported impact
The targeted account experienced a cliff-like drop in post traffic after the coordinated attacks. The blogger noted that the attack method creates psychological pressure through unexplained traffic loss, causing self-doubt and creative burnout.
Commentary source文昭 stated that since mid-2025, many content creators on X and YouTube have experienced traffic anomalies attributed to similar operations. The commenter described the effect as 'choking propagation and economic attack'.
Index analysis
The attack exploits a gap in platform governance: unfollowing and blocking are user rights, making them difficult for platforms to classify as attack behaviour, providing 'natural compliance cover' for operators.
The commentator文昭 indicated that Chinese state actors have systematically studied major platform algorithms for over two years, using AI tools to lower the operational threshold and cost of large-scale manipulation. Another observer noted coordination between water army accounts and professional technical teams, citing a case where a blogger with 80,000 followers was permanently banned after false reports by coordinated accounts.
Governance controls
The targeted blogger recommended periodic manual deletion of blank accounts to reduce algorithmic association with low-quality graphs, and increased posting and engagement during attacks to counter negative signals with positive interactions from genuine followers.
Commentator文昭 suggested immediate cleaning of comment sections and blocking suspicious accounts to prevent coordination and avoid triggering platform suppression mechanisms.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.