Facebook announced Monday the deployment of an artificial intelligence system to identify more quickly social network members prey to suicidal ideation to help them.
The technology will automatically look for clues in messages as well as videos uploaded live on Facebook Live, to alert Facebook staff and specialized help organizations more quickly, according to the US group.
“This approach uses behavioral pattern recognition technology to help identify live messages and videos that potentially express suicidal thoughts,” wrote Facebook vice president of product management, Guy, in a blog post. Rosen.
The artificial intelligence system will sift through the messages of members but also the responses of their friends who could for example ask them if all is well or “what can I do to help you?”, Explains Guy Rosen.
Facebook has already put tools in place that allow users to report if a friend seems to want to end their day, but artificial intelligence makes it possible to report this information more quickly and even detect signals that members may not be able to detect. .
“Terrible tragic events – including suicides, some broadcast live – may have been avoided if someone had realized what was going on and had made it known earlier,” said the boss. Facebook, Mark Zuckerberg, in February.
“Artificial intelligence can help provide a better approach,” he said in an open letter.
After testing it in the United States, Facebook will deploy this artificial intelligence system outside the United States. “It will eventually be available anywhere in the world, except in the EU,” said Guy Rosen without further details. Personal data is protected in the European Union by stricter regulations than in the United States.
Facebook has been working with mental health organizations for the past decade to combat suicides and help its distressed members find the support they need, says Rosen in his message.