Three weeks before Americans vote in the November 3 midterms, X says it removed more than 13,000 accounts it suspects were part of foreign influence operations linked to Iran, Russia, and China. The announcement is significant, but what X left out of its public statement is equally important.
The company’s public materials included no technical evidence, no account breakdown by country, and no independent verification of its attribution. Readers are left to weigh a serious allegation on X’s word alone.
The Takedown
X says its Safety team identified more than 13,000 accounts through proactive sweeps, but the public record stops well short of explaining how those accounts were traced to three countries.
X’s Safety team says it identified the accounts through proactive sweeps targeting coordinated inauthentic behavior and spam networks. According to the company, the accounts used inauthentic identities to post and amplify political content involving US candidates and public figures across the political spectrum.
X did not disclose how many accounts were linked to each of the three named countries. The 13,000 figure represents the combined total across all three suspected operations, as reported by Al Jazeera and Khaama Press.
What X Did Not Say
X named three countries but provided no technical indicators, infrastructure links, or intelligence assessments in its announcement to support those attributions.
X publicly attributed the suspected networks to Iran, Russia, and China, but the announcement did not disclose technical indicators, coordinated-posting patterns, or infrastructure evidence to support those country-level links. The phrase “suspected foreign influence operations” is X’s own characterization, not an independently established conclusion.
The sources reviewed for this article did not document a response from Iranian, Russian, or Chinese officials to X’s specific claims. That evidentiary gap matters. Outside researchers, journalists, and civil-society organizations may be unable to independently verify whether the attributed accounts connect to the countries named, based on the public materials available. A surveillance app built to target political dissidents illustrates how covert digital influence operations extend well beyond any single platform or border.
MI5 Director General Sir Ken McCallum has publicly documented Russian interference attempts around the 2016 Brexit referendum, including activity involving online trolls and bots. He cautioned, as reported by Reuters and the BBC, that attempted interference should not be treated as proof that an electoral result was changed.
That distinction applies directly here. A removal action is not, by itself, confirmation that a foreign government successfully manipulated political discourse.
Platform Safeguards and Their Limits
X’s election-integrity tools include Civic Integrity reporting and Community Notes, but both face unresolved questions about reliability at scale.
X’s Civic Integrity policy covers misleading claims about how, when, or where people should vote, as well as content intended to discourage participation. Users can flag election-related posts through a dedicated reporting category, with submissions routed into requests for Community Notes. In critical cases, X says it may also apply a Civic Integrity label directly.
Those tools face documented scrutiny. When the European Commission opened formal proceedings against X under the Digital Services Act in 2023, it raised specific questions about whether Community Notes can reliably assess post accuracy. Concerns about platforms secretly tracking users add another layer to the broader accountability questions surrounding digital services at scale. The gap between proactive account removal and crowdsourced moderation that depends on volunteer consensus remains unresolved heading into Election Day.
X’s choices on attribution transparency, account removals, and labeling will face continued scrutiny through November 3 and beyond. Knowing that 13,000 accounts are gone is a starting point, but democratic accountability requires understanding why they were removed and how those attributions were made. A detailed technical explanation is not included in the public materials cited here.
