On Wednesday morning local time, some X and Reddit users discovered that searching for the name of OpenAI CEO Sam Altman on Google resulted in a search result indicating that he had died that day. A few hours later, Google admitted that the search result was incorrect and responded on X: "The information is no longer displayed. This was not a manual modification by Google. When someone maliciously alters public information sources, it may affect the information displayed in the search results."

image.png

Netizens generally speculated that this was related to the malicious tampering of Wikipedia. The revision history of Sam Altman's Wikipedia page showed unusually frequent activity on that day, with 22 edits made in just one day, while the page had only been updated five times since the beginning of the month, and each update occurred on different dates.

"Death Announcement" to "Name Change": A Malicious Operation Lasting Several Hours

The timeline of the incident reveals the farce: at 9:47 AM Coordinated Universal Time, the page was modified to show false information stating that Altman was assassinated on Wednesday; at 10:29 AM, the content was corrected. However, the page then underwent a series of modifications, including adding a description that Altman was a "dropout," reasserting his death, and even maliciously changing his surname from "Altman" to "Gayman." As of the time of this report, the latest update seems to have corrected all previous errors, but this malicious editing still exposed a vulnerable aspect of the search ecosystem.

The core controversy of this incident lies in Google's Knowledge Panel. As an information area displayed at the top of some search results, the content of the Knowledge Panel comes from various public information sources, with Wikipedia being one of the most frequently cited. Google previously explained that its information comes from hundreds of sources across the internet, not solely relying on Wikipedia, and has an information reporting mechanism where users can flag incorrect information, and once policy violations are identified, "it will be manually removed." However, this incident shows that in the AI era, increasingly reliant on real-time information retrieval, a publicly edited encyclopedia page being maliciously rewritten is sufficient to contaminate search results and create large-scale rumors within a short period of time. How to build a stronger defense between the massive amount of public information and content credibility remains a long-term challenge for platforms like Google.