Google Search falsely declared Sam Altman dead, but Google says nobody did it manually

Google Search briefly advised customers that OpenAI CEO Sam Altman had died on August 12, 2026, displaying the false date prominently inside his Information Panel:

Google has now responded to the backlash and eliminated the misinformation.

In a public response on X, the corporate mentioned the end result was not manually modified by anybody at Google and blamed vandalism affecting a public data supply.

How Google Search by chance killed Sam Altman

In our exams, Search Engine Watch discovered that Sam Altman’s Wikipedia biography was changed so as to add a dying date of August 12, 2026, a dying location in Seattle, and an unsupported assassination declare.

The opening paragraph was additionally become the previous tense, and Wikipedia robotically tagged the change as a doable unreferenced addition to a biography of a dwelling particular person.

One other editor reverted it at 10:29 UTC, which means the false data remained on Wikipedia for about 41 minutes. However 41 minutes was greater than sufficient for Google to replace its search outcomes with Sam Altman’s dying date earlier than it was quietly rolled again.

At 12:36 UTC, an X person published a screenshot exhibiting Google’s Information Panel itemizing Altman as having died that day. One other broadly shared screenshot adopted minutes later.

A second non permanent Wikipedia account reintroduced the false date at 12:39 UTC earlier than it was once more eliminated a number of minutes later.

After additional disruptive enhancing, Wikipedia positioned the web page below extra safety.

The publicly out there timestamps can not set up exactly when Google first imported the false data or precisely how lengthy it remained seen. Screenshots could be captured earlier than they’re posted.

Nonetheless, the sequence strongly factors to Wikipedia because the supply and reveals how simply a risky edit entered, or remained inside, considered one of Google’s most authoritative-looking Search options.

This was not an AI Mode hallucination

The error appeared in a traditional Information Panel below Google’s normal “All” outcomes tab. It was not generated by AI Mode or an AI Overview.

This seems to have been a structured information and Information Graph failure, though automated extraction was nonetheless accountable for turning the unhealthy supply data right into a direct reply.

The panel additionally contradicted itself. It displayed a big “Died” area whereas persevering with to explain Altman as the present “CEO of OpenAI.” Its Wikipedia-based overview used the current tense and mentioned he “is an American entrepreneur.”

Google says it didn’t edit the search outcomes manually

Google responded:

“Thanks for flagging. That is now not showing, and was not a guide change by Google.”

The corporate added that vandalism of public data sources can have an effect on what seems in Search.

That assertion is factually cheap, however it solutions the least tough a part of the issue. There is no such thing as a proof {that a} Google worker manually altered the panel, and critical criticism doesn’t require such a idea.

Google’s personal documentation says Knowledge Panels collect information from various sources and replace robotically as data modifications throughout the online. Its Information Graph documentation equally says details come from a variety of public, licensed, and first-party sources.

If a single lately edited Wikipedia web page was enough to ascertain {that a} distinguished dwelling particular person had died, what corroboration occurred? If Google consulted a number of sources, why did the absence of any supporting reporting not forestall the sphere from showing? Why was an robotically flagged Wikipedia edit handled as sufficiently dependable for such a consequential declare?