X says it has caught a foreign hand stirring one of America's loudest tech fights. On Friday, the platform's Safety team said it had uncovered roughly 200,000 suspected inauthentic Chinese accounts — and that about 200 of them were posting specifically about AI data centres: comic strips, images and text arguing that the boom drives up household electricity bills and strains local grids, including AI-generated cartoons of data-centre operators getting rich while ordinary families pay.

X framed it as serious foreign interference. The posts, it said, "could manipulate a legitimate debate about American AI and energy policy," adding: "We take seriously any attempts to undermine the integrity of the global town square." There's an irony baked in: according to Engadget's account, the bot accounts leaned on OpenAI's ChatGPT to generate their illustrations and talking points — AI used to argue against the buildout of AI.

Two things can be true at once

The disclosure landed in a debate that was already hot, and it cuts two ways.

First, the caveat X did not lead with: the grievances the bots amplified are real. US electricity prices rose about 6.9% year over year in 2025, by a Goldman Sachs analysis published in February 2026, and communities next to hyperscale construction have spent roughly two years raising serious complaints about rates, water use and noise — with some towns rejecting data-centre proposals outright. A bot farm boosting a message does not make the message false. The households seeing higher bills are not Chinese accounts.

Second, the messenger has skin in the game. X is owned by Elon Musk, who is himself building enormous AI data centres — which is exactly why critics greeted the announcement warily, reading a claim that "opposition to data centres is a foreign psy-op" as conveniently self-serving for a man who wants more of them approved. X has not published the granular evidence behind the attribution, so the public is being asked to take the Safety team's word for the scale and the intent.

Why it matters

This is the information-integrity layer of the datacentre fight, and it's badly tangled: a foreign influence operation, a platform owner with a direct commercial interest in the outcome, and a wave of real public anger all occupying the same comment threads. The risk is that a true finding — that some accounts were inauthentic — gets stretched into a false conclusion, that the opposition itself is manufactured. It isn't. The honest reading is narrower and more uncomfortable: bad actors are real, the platform flagging them is conflicted, and the underlying complaint about power bills and water was never theirs to begin with.