Paris: The International Olympic Committee is using artificial intelligence to identify and block online hate aimed at 15,000 athletes, delegates, and officials taking part in the Paris Games, in a world first for the landmark sporting event.
In the first week of the competition alone, organisers have successfully lobbied social media platforms to remove more than 1000 posts identified using the technology, which can scour open-source sites in more than 37 languages.
Some 15,000 athletes, delegates and officials have assembled in Paris.Credit: Michel Euler
IOC head of safe sport Kirsty Burrows said organisers anticipated about half a billion social media posts during the Olympic and Paralympic Games in Paris, which made monitoring them by hand impossible.
“If you would take one second per post it would take you 16 years nonstop to go through them,” Burrows said.
“So it was about the speed first of all, and also about being able to better understand it because we’ve always looked to try and improve safeguarding measures.”
Unlike human beings, the technology scans millions of data points online in real-time to find hateful posts and comments. It then sorts them into categories based on the nature of the abuse and the risk it poses using an algorithm.
That information is then forwarded to a team of human experts, who check the nature of the posts, report serious concerns to the police, and contact social media platforms to request that they take the material offline.
“We then action safeguarding support on the ground for anyone who has been targeted,” said Burrows, although the idea is to get the content removed before athletes can even spot it.
