The star platform had to beef up its game against insulting tackles and virtual harassment.
Instagram recently boosted its ability to moderate comments and private messages aimed at football players, who are often exposed to public condemnation at competitions such as the World Cup in Qatar.
"We must protect an online space where we unfortunately find the same behavior as in the stands of matches," laments Tara Hopkins, director of public policy at Instagram.
The subsidiary of Meta (ex-Facebook) has therefore recruited a dedicated team from the sports world in order to forge direct links with footballers, clubs and agents.
This group of multilingual moderators was first mobilized for the last Olympic Games.
Insta then invited all the players to a meeting at the end of October to remind them before the World Cup of the tools developed to counter the surge of insulting comments or private messages.
A phenomenon amplified by the media exposure of footballers.
Emojis diverted to insult
This online hatred has taken all forms year after year: racist and homophobic insults but also emojis diverted from their use.
“We observed a spike in malicious activity when a goal was scored or missed,” notes Tara Hopkins.
The algorithms, which filter the content, have been fed with artificial intelligence and in case of doubt a human check has been added to decide on the illegality of an anonymous comment or not.
It becomes crucial to know the context in order to intervene: during the Euro 2020 final where black players were targeted after the defeat of England by a campaign of racism with banana emojis, monkeys or palm trees.
Read alsoFaced with criticism, Instagram takes a break from its TikTok transformation
The social network also cleans up and anticipates offensive remarks in “DMs”, private messages.
A player or his community manager can opt for increased filtering with the “Hidden words” tool.
The artificial intelligence will then automatically hide hurtful terms in a message or in public comments.
This content lands in a separate message box that is searchable and easy to report to the platform.
“1 account with more than 10,000 subscribers out of 5 has activated this option for a year, it corresponded well to a need”, comments Tara Hopkins.
Holes in the racket
Another tool available to high-level athletes: the possibility of limiting interactions with people who do not follow them.
Once again, the AI comes into play and detects - sometimes before the wave hits - a sudden mass of comments or private messages and suggests limiting this influx.
“Without any risk of saturating the filters because they were designed in collaboration with players and are adapted to absorb peaks of activity, assures the manager.
The net is not 100% perfect because everything is often a question of context in harassment, ”she admits.
Read also Criticized from all sides, TikTok strikes back
Instagram distributes a few warnings in the form of temporary suspension but refuses to give the conditions of the red card, the deactivation of an account, "so as not to encourage users to push the limits of the rules".
The competition, like TikTok, has also been proactive with an equivalent automated hate comment filtering system.
A special system for the World Cup has been put in place "to strengthen benevolence on our platform".
“We are going to add messages of public interest on hashtags related to football, explains TikTok France.
These posts will invite our community to explore our Safety Center and remind them of the importance of reporting inappropriate content or behavior.”
The online hate game is far from over.
