In May 2026, posts on Facebook claimed the historian Heather Cox Richardson started a livestream at 3 a.m. to announce she had received a threatening message from a powerful politician. While the posts never stated who the politician was directly, they were accompanied by images suggesting it was U.S. President Donald Trump.
The claim has been spreading since at least April 16, when a
The posts began:
HEATHER COX RICHARDSON GOES LIVE AT 3 A.M. WITH AN URGENT MESSAGE
"I received a message tonight — and it was meant to silence me."
Washington, 3:07 a.m. — Heather Cox Richardson did not wait for a press conference, did not release an official statement, and did not choose a carefully prepared address. Instead, she unexpectedly went live in the middle of the night. No formal setup. No staff behind the camera. No applause.
Snopes readers wrote in asking for more information about the supposed threat against Richardson and her response.
We found no information to support it.
Searches of DuckDuckGo, Google and Yahoo found no news media outlets reporting about the rumor, nor evidence from Richardson's social media accounts. The claim instead originated from Facebook accounts and blog pages that use artificial intelligence tools to create inspiring or shocking stories about public figures. Therefore, we've rated this claim false.
Snopes has checked three nearly identical claims about Pope Leo XIV, comedian Stephen Colbert and figure skater Ilia Malinin. The fact-checking website Lead Stories published an article in December 2025 documenting examples of the claim targeting 30 different people.
Creators of such content capitalize on social media users' willingness to believe and share the made-up stories, profiting from advertising revenue on external websites to which the posts link. (We have previously reported on this business strategy.)
The posts are often accompanied by a link to an unreliable blog in the comments, which generally post a massive number of articles, with vague descriptions and dramatic wording typical of AI-generated content. The blogs often substitute a few characters of the Latin alphabet for non-Latin characters (generally "n" and "u" for the Cyrillic "п" and Greek "υ"), which is likely an attempt to avoid moderation or ad restrictions.
In order to confirm that the text is AI-generated, Snopes usually runs the text of such articles and posts through GPTZero, an AI-detection tool. These types of tools are fallible, however, and Snopes cautions people against using them for definitive answers on media's authenticity without supporting evidence.
The blogs had disabled the ability to highlight body text, so we manually typed the first five paragraphs into GPTZero. We also checked the Facebook posts' caption. The tool determined with 100% certainty that both the article and Facebook posts' captions were AI-generated.
Our suspicions were confirmed by looking at the Lil chase Facebook page, which revealed an inhuman posting volume and images that were clearly AI-generated. Snopes contacted a manager of the Lil chase Facebook page to ask why it had posted the false story about Richardson without including a disclaimer, and will update this story if we receive more information.