On X, they post as The General. The account, which comes with a blue check mark and describes itself as “Constitutionalist, Patriot, Ethnically American,” has been sharing a steady stream of pro-Trump, U.S.-centric content since March of 2016.
But The General may not be what their followers envision: The account is actually based in Turkey.
It’s just one of many highly active and relentlessly partisan accounts on X, the social media platform owned by Elon Musk, that have been affected by a new feature allowing users to see posters’ country of origin. This geolocation data has unmasked scores of accounts that generate millions of impressions posting about U.S. politics and other hot-button issues despite being located outside the U.S. The transparency feature, released on Nov. 21, has itself become a viral topic on X and other platforms.
Why We Wrote This
A new feature on the social media platform X is revealing that many popular accounts featuring inflammatory content about U.S. politics are located in Asia, Africa, and Eastern Europe. While it’s hard to know who is behind them, experts say many are just trying to profit from outrage.
While the geolocation data is incomplete and, in some cases, disputed by account holders, users have seized on revelations that some prominent X accounts appear to be frauds. Pro-Trump posters who extol the virtues of America First have turned out to be based in Nigeria and Bangladesh. Progressive content has been amplified by foreign posters who purport to be U.S. voters.
Who is behind these accounts isn’t always clear. Foreign governments have been known to use social media to spread disinformation and influence U.S. elections. Among the best documented of these efforts was Russia’s attempt to stoke tensions over police shootings and Black Lives Matter protests in 2016, using fake personas on both sides.
But researchers who study misinformation say much of the divisive content is motivated by something more basic: money. Posters on X can profit from clicks. And what reliably engages many U.S. users is outrage-inducing political content that can be wildly inaccurate and still deemed credible by partisans.
Still, as more and more Americans derive their news from social media rather than mainstream journalism sources, the impact of that grift can be far-reaching.
“The core issue here is the inherent unreliability of the information ecosystem,” says Mor Naaman, a professor of information science at Cornell Tech in New York City. “Our information ecosystem is built on incentives that very much invite anyone to post divisive or other engagement-based content, because sometimes they make money.”
In the Middle East, X’s geolocation data shows some pro-Palestinian posters who purport to be in Gaza are located elsewhere, raising doubts about their authenticity. Some accounts that relay images of life under duress in Gaza and ask for donations have been exposed as potential scams. In the United Kingdom, some accounts that advocate for Scotland’s independence appear to originate in Iran and the Netherlands.
In the U.S., this “great unmasking” has shone a spotlight on a slew of “patriotic” MAGA accounts that are located in Asia, Africa, and Eastern Europe. Many of these accounts feature AI-generated images of blond-haired American women and of President Donald Trump and his family. Surprising locations have also turned up for left-leaning accounts – such as one called Republicans Against Trump, with nearly half a million followers, which was shown to be based in Austria before changing its location to the U.S.
Analysts say incentives for “rage baiters” are baked into X’s business model under Mr. Musk’s ownership. Take the blue check mark, which was originally meant to signal that an account holder was who they claimed to be. It has now become a pay-to-play badge that helps content makers, authentic or not, build audiences that can be monetized via revenue sharing with the platform.
“That makes it easier, not harder, for fake or proxy accounts to appear credible,” says Emilio Ferrara, a professor of computer science at the University of Southern California, via email.
At the same time, X has cut its moderation and trust-and-safety teams and instead asked users to provide context for misleading posts in the form of community notes. Crowd-sourced tools can be helpful, but it’s hard to assess what X’s features mean for fake and manipulative content since X no longer allows independent researchers to access its data, says Professor Ferrara.
Some accounts have responded to X’s geolocation data by claiming to be Americans living or working abroad. The General, for example, posted a video of a U.S. passport and claimed to be in Turkey on a business trip.
When announcing X’s new geolocation feature, Nikita Bier, the company’s head of product, called it “an important first step” and promised to introduce more ways for users to verify the authenticity of content. In a later post, he said it still had “a few rough edges” to be resolved. Some account owners have complained about inaccuracies and said the use of VPNs and other proxy connections could lead to mislabeling of legitimate accounts.
There are technical challenges in identifying what is and isn’t an authentic “grassroots” voice in online debates, since foreign actors can be sophisticated at hiding their tracks, says Professor Ferrara, who studies inauthentic behavior on digital platforms. “People may put too much trust in a signal that is still inaccurate,” he says of X’s new feature.
In a recent Washington Post op-ed, Zach Schapira, a former X executive, and Sean Rad, founder of the dating app Tinder, call for X and other platforms to “expand geographical transparency” by showing users where viral posts are being generated and spread, and giving them options to filter from specific regions. They also recommend that algorithms prioritize authentic domestic content. “Americans debating issues that affect their communities deserve a feed that reflects real local voices and the real local popularity of those voices,” they write.
Still, some worry that any whack-a-mole approach will fail, as duplicitous posters rely on VPNs and other tools to mask their origin while continuing to use AI tools to pump out content.
The targeting of Americans, in particular, with provocative and divisive political content is, in part, a reflection of U.S. consumers’ spending power. Popular MAGA content frequently attracts ads for gold and male health supplements, among other products. “It could be that there are just more immediate products to sell to a right-leaning” X user, says Professor Naaman. (The platform also skews right; Bluesky, another social-media platform, has become a liberal hub.)
Mr. Musk, who has 229 million followers on X with whom he shares a daily diet of right-wing political content, hasn’t yet commented on the geolocation revelations. It’s something of a double-edged sword for his stewardship of a platform that he frames as the world’s best news source, says Renée DiResta, an associate research professor at Georgetown University. The added transparency “helps people be more aware that these kinds of rage baiters exist in their information ecosystem,” she says. At the same time, it’s highlighted “that the platform does have these perverse incentives” for inauthentic posters to make money.
Even if account holders aren’t disguising their country of origin, “being confrontational and provocative and divisive” reaps rewards, says Professor DiResta. “Everybody in tech knows this is happening.”
