A man collapsed in the Oval Office. The internet got the story wrong.

A case study in how bad information can spread — and how hard it is to correct it.

There was a scary moment at the White House on Thursday, when a man standing behind President Trump suddenly collapsed during a drug-price event.

You may have seen photos or videos — it was one of those scenes that went viral very quickly.

I was covering the event for The Washington Post, watching as Trump officials and pharmaceutical executives from Eli Lilly and Novo Nordisk touted their new deal to lower the price of GLP-1 drugs. Dave Ricks, the CEO of Eli Lilly, had just started speaking when the man began to lose his balance.

“Gordon, you okay?” Ricks asked the man as he struggled to stand and fell to the floor.

White House officials abruptly ushered people out of the room and halted the event. As questions swirled, some reporters and popular social media accounts quickly identified the man as Novo Nordisk executive Gordon Findlay. I don’t know how they did it — perhaps it was as simple as googling “Gordon” and “Novo Nordisk,” and looking for a resemblance.

Within minutes, “Gordon Findlay” was trending across social media. Aggregated articles rapidly appeared, purporting to explain who “Gordon Findlay” was and what had happened at the White House. And as I write this post, hours later, summaries of the drug-price event still mention the “Novo Nordisk Executive” who collapsed in the Oval Office.

Just a few problems with that.

First, the man was not “Gordon Findlay,” who as far as I can tell, appears to be based in Switzerland and does not hold a senior Novo Nordisk position that would score him a coveted invitation to the Oval Office.

Second, the man was not a pharmaceutical executive — but an Eli Lilly patient, who had been invited to the White House based on his experience taking a GLP-1 drug.

How did I know this? I did some rapid reporting: I talked to people with direct knowledge of the White House event and learned who was invited. I got the name of the man who collapsed. (Which, yes, begins with “Gordon,” but doesn’t end with “Findlay.”) I saw photos of him in normal life that clearly matched the man at the White House.

So I posted on X and Bluesky that, no, the man was not “Gordon Findlay” nor a Novo Nordisk executive but an Eli Lilly patient. I noticed a few other reporters soon working to correct the record, especially after Novo Nordisk said that “Gordon Findlay” wasn’t in attendance. Several hours later, Eli Lilly confirmed that the man who collapsed had been their guest (though didn’t share further details).

I thought that would be the end of the story, as accurate information slowly trickled out and supplanted the misleading claims.

What I wasn’t expecting — the sheer number of people who continued to argue with me, convinced that “Gordon Findlay of Novo Nordisk” had collapsed at the White House, based on their comparison of a few photos. I don’t want to link to those arguments and amplify them, but if you go onto X and Bluesky, you can easily find several of my responses.

Even now, hours later, many prominent social media accounts — boasting blue checks on X and with hundreds of thousands of followers — have yet to delete their posts that wrongly claim the man who collapsed was “Gordon Findlay.” Grok, Elon Musk’s AI tool, continues to falsely identify the man when asked to explain who collapsed at the White House.

And I’m still hearing from social-media users who assert, confidently, that Gordon Findlay fell to the floor of the Oval Office on Thursday.

Ultimately, the damage from this mis-identification may be small. I’m sure the real Gordon Findlay has had a very strange day. Meanwhile, the Gordon who collapsed in the Oval Office swiftly recovered, officials said.

But to me, it’s a telling reminder of the perils as social media, AI, and aggregation increasingly drive our news consumption. It shows how internet users can quickly develop and then defend false beliefs — despite being offered better information.

And the inability to correct something this small and specific doesn’t bode well for the bigger news-cycle missteps we’ll inevitably face.