Following the Notre Dame fire, multiple media outlets broadcast live coverage of the event through YouTube. However, the algorithm of the social network seems to have confused this event and has thrown an annotation of the Encyclopedia Britannica on September 11, the day of the fall of the twin towers.
“This panel with data on the fall of the twin towers has only been identified in the news direct CBS, NBC news and France 24h.”
The error arose in what YouTube calls its “knowledge panels.” These panels are part of the advances of the platform to protect against misinformation with data verification processes. In this sense, the panels are designed as a response of the algorithm to provide more information in videos about sensitive topics or current events that become very popular.
In the case of the direct ones of the cathedral of Notre Dame, the algorithm recognized the images of the fire and its popularity in the app and automatically related it to other events recorded in the database of these panels, which works associated with pages such as Wikipedia and the Britannica Encyclopedia.
The objective of the algorithm is to provide verified information to avoid the rise of conspiracy theories. However, by relating the terrorist attack of 9/11 with an accidental fire, the platform ends up spreading more disinformation than it planned to eliminate.
Fortunately, this panel with data on the fall of the twin towers has only been identified in the news direct CBS, NBC news and France 24, so the error has not had a massive diffusion. Either way, YouTube authorities have taken action on the matter and temporarily disabled the knowledge panels for any image related to the fire.
However, this error in the system has generated many criticisms about how algorithms are handled for social networks. In fact, several experts have taken the event as a sign that computerized learning systems and artificial intelligence must be supported by the review of a human moderator to ensure its effectiveness.