China is weaponizing ChatGPT to spread anti-AI propaganda aimed at Americans, according to OpenAI.

In a report published Wednesday, the company said it shut down a fleet of bogus ChatGPT accounts generating content for social media posts that depicted data centers as being responsible for American households’ rising energy costs. The creators of the accounts used VPNs (ChatGPT isn’t available in China) and appear to have been “conducting work for Chinese provincial-level government clients,” according to OpenAI.

The AI-powered disinformation campaign’s reach and impact were minimal, the report said. Still, it was a glimpse of a future in which foreign adversaries harness publicly available AI tools to exacerbate political divisions within the U.S. to gain a geopolitical edge. “The operation sought to exploit and amplify existing public concerns about energy prices and local impacts of data center development,” the report said, “but we found no evidence of meaningful breakout beyond its own activity.”

In addition to the accounts that were trying to fan the flames of anti-data center sentiments, another group of now-deactivated accounts was being used to generate social media posts portraying the Trump administration’s Tariff policies as stifling tech competition abroad.

A political boon

OpenAI’s evidence of government-backed Chinese hackers trying to oppose the American AI industry could be welcome news for many Republicans.

Earlier this month, GOP lawmakers began calling upon the Trump administration to investigate “foreign influence campaigns targeting artificial intelligence (AI) development in the U.S.,” according to an open letter addressed to David Sacks and Michael Kratsios, co-chairs of the President’s Council of Advisors on Science and Technology, and FBI Director Kash Patel. 

The letter cited a report published last month by the Bitcoin Policy Institute, a think tank “dedicated to advancing sound Bitcoin policy,” according to its website, which pegged the blame on what it described as a “campaign against American AI” on a triumvirate of Chinese state media, a network of left-leaning organizations funded by American tech entrepreneur Roy Singham, and “foreign billionaire dark money.” Also cited in the letter was a report from Power the Future, a right-leaning energy industry advocacy group, which claimed that leftist “environmental activists” have been channeling money provided by ideologically motivated billionaires through nonprofits to stymy the growth of new data centers, “creating the false appearance of grassroots resistance while advancing a broader anti-growth, anti-Trump agenda.”

Republicans have generally adopted a laissez-faire attitude towards the expansion of domestic AI infrastructure, widely supported by the belief that a slowdown at home would advantage China. On Wednesday. Environmental Protection Agency (EPA) administrator Lee Zeldin said the agency would not step in to establish nationwide regulations around the spread of data centers, and that such decisions were better left to local governments.

Increasingly power-needy tech companies have therefore been given ample leeway to build new data centers, sparking pushback from communities, environmental advocacy groups, and state lawmakers across the country.

The new report from OpenAI—one of the most well-known (and until very recently the most well-funded) AI labs in the world—is a considerable boon for Republicans who are trying to blame anti-AI pushback at home on shadowy foreign actors. It’s also a convenient narrative for tech developers themselves as they seek to expand power supplies in the face of mounting community opposition: If your opposition is rooted in a misinformation campaign intended to subvert American interests at home and abroad, then surely you have a right—a patriotic duty, even—to carry on full steam ahead… right?

Foreign bogeymen

There is, of course, a long history of state-backed hackers using new technology to sway American public opinion and undermine democratic institutions, such as elections. This is already being supercharged by AI, as OpenAI’s new report makes clear. And as the company rightly points out, tech companies building AI and government agencies, therefore, need to be on their guard.

But it would be at least as dangerous to let fears of ideological bogeymen eclipse the fact that communities across the country have real anxieties about the encroachments and energy demands of AI data centers. Nefarious hackers will almost certainly continue to exploit those concerns, but that doesn’t delegitimize the concerns themselves.