—Travis Luyindama, B1Daily

China-linked influence operations are no longer confined to clunky propaganda videos praising Beijing. Investigators say some networks have learned a potentially more effective trick: pretend to be Americans, exploit the country’s political and racial fault lines, and use artificial intelligence to manufacture content at industrial speed.

The evidence requires an important distinction. There is substantial documentation of PRC-linked influence operations using AI and exploiting racial tensions, including activity on TikTok. But available evidence does not establish that every racist AI video circulating on TikTok is produced or directed by the Chinese government.

What is documented should still concern Americans.

TikTok Becomes Another Battlefield

The Chinese-linked network commonly called Spamouflage or Dragonbridge has operated fake social-media personas pretending to be Americans. Graphika identified accounts discussing racial inequality, gun control, homelessness, drugs and other divisive U.S. issues. Some of the political material was assessed as almost certainly AI-generated. One identified TikTok account managed to draw 1.5 million views before TikTok banned it.

The operation wasn’t simply promoting one American party. Researchers found content attacking figures and narratives across the political spectrum, apparently seeking to deepen distrust and division rather than reliably elect one side. China has denied interfering in U.S. elections.

That distinction matters. Foreign influence does not necessarily require convincing Americans to become pro-China. Making Americans more suspicious of one another can itself serve a geopolitical purpose.

Race Is an Obvious Pressure Point

Microsoft researchers found China-linked accounts concentrating on contentious American subjects including racial tensions, immigration, border policy, drugs and other domestic disputes. Researchers also observed accounts posing questions apparently designed to learn which political subjects generated the strongest reactions among different groups of American voters.

Meanwhile, researchers have documented a separate explosion of grotesque AI-generated racial material inside China’s own social-media ecosystem. Foreign Policy reported in 2025 that fabricated videos depicting interracial relationships and other scenarios were being used as rage bait, with anti-Black content particularly prevalent on Chinese platforms including Douyin, China’s domestic counterpart to TikTok. That reporting demonstrates the availability and popularity of this material, although it does not establish that the Chinese government created those particular videos.

Put the developments together and the danger becomes clearer: cheap generative AI makes it possible to manufacture provocative racial imagery, fake personalities and political messaging faster than yesterday’s troll farms could dream of doing.

This Is Bigger Than TikTok

The operation is cross-platform. Reuters reported that Spamouflage has operated across more than 50 websites, forums and social-media platforms, while Graphika found personas pretending to be American voters and activists.

U.S. intelligence has also warned that PRC actors are improving their ability to conduct covert influence operations. The Office of the Director of National Intelligence reported that pro-China actors used AI-generated news anchors and fake accounts with AI-generated profile images while attempting to sow division around immigration, abortion, drugs and other American controversies.

OpenAI separately disrupted accounts associated with Spamouflage. The operation has been attributed by Meta to individuals associated with Chinese law enforcement, while some earlier activity was attributed by the FBI to a unit within China’s Ministry of Public Security. OpenAI found the network using AI for research and content generation, although it also concluded that AI-generated material represented only a minority of the operation’s overall output at the time.

America’s Racial Problems Shouldn’t Become Foreign Propaganda Ammunition

None of this means America’s racial problems are inventions from Beijing. Racial discrimination, political polarization and inequality are American issues that Americans have every right to debate fiercely.

The problem begins when a foreign influence operation masquerades as Americans and secretly inserts itself into that conversation.

A fake account pretending to be a Black voter, conservative activist, progressive organizer or frustrated working-class American isn’t participating in democratic debate. It is counterfeiting participation.

AI makes that counterfeit cheaper. TikTok and other social platforms give it distribution. America’s existing divisions provide the raw material.

The proper response isn’t censorship of legitimate arguments about race. It is transparency: stronger identification of coordinated foreign influence networks, aggressive removal of deceptive state-linked accounts, disclosure of synthetic political media, and greater scrutiny of who is actually behind accounts claiming to speak for American communities.

America’s political arguments should belong to Americans, not anonymous operators thousands of miles away wearing AI-generated American faces.

—Travis Luyindama, B1Daily