State-affiliated influence campaigns from Russia and China have entered a new phase, marked by the sophisticated use of artificial intelligence. These operations now focus on crafting online personas that closely mimic human behavior, a significant evolution from their previous tactics that relied on sheer volume and automated bots.
The Evolution of Influence Tactics
Historically, researchers tracked numerous inauthentic accounts promoting pro-Russian and pro-Chinese narratives across platforms like X, formerly known as Twitter. These accounts were characterized by high-frequency posts produced by automated systems. However, the strategy has shifted to involve fewer posts, richer content, and human-like activity patterns, including mimicking sleep schedules to evade detection.
Analysts at TwoSixTech have developed a machine learning model to effectively identify these accounts on X, analyzing data from the years 2024 through 2026. The analysis revealed a reduction in post volume by half, with a noticeable improvement in content quality, as disclosed in a report to Cyber Security News.
Strategic Use of AI and Older Accounts
The findings indicate a strategic learning curve among these actors. Instead of creating new profiles, they now repurpose existing accounts for new campaigns. A notable example includes a pro-Kremlin campaign targeting Armenia’s prime minister, Nikol Pashinyan, during elections, as reported by TwoSixTech.
AI-generated images and multilingual content are increasingly utilized, signaling a shift towards influencing public opinion with more engaging and persuasive material. The number of active inauthentic accounts on X remains substantial, with thousands attributed to both Russian and Chinese operations.
Targeting the United States
In a significant shift, pro-Russian actors have escalated their attacks on the United States and its leadership, including President Trump. This change underscores Moscow’s dissatisfaction with the administration’s stance on Ukraine. Between 2024 and 2026, anti-US narratives surged, with a 264% increase in personal attacks on US figures and a 263% rise in messages about military weaknesses.
Pro-China actors have consistently promoted anti-US narratives, positioning China as a global leader in AI. In 2026, the campaign expanded to portray Japan’s Prime Minister Sanae Takaichi as a US puppet, broadening their geopolitical focus.
TwoSixTech advises social media platforms to enhance their AI-driven detection capabilities, focusing on behavioral patterns rather than mere post frequency. Monitoring the reuse of older accounts is also recommended to thwart these increasingly sophisticated influence networks.
