contador javascript Skip to content
Contact :

Against bullying: Instagram detect offensive texts with artificial intelligence

The application automatically detect potentially offensive texts and issue a warning for the user to evaluate the publication of the content

In a measure that seeks to limit online harassment,
 He announced that he began using artificial intelligence systems to detect offensive texts when publishing a photo or video. "In this way we give people an option to reconsider their words before publishing them," the company said in a
official statement

Previously, Instagram had implemented a similar system to identify and notify the authors that
posted offensive comments. "The experience has been very good and we discovered that these small warnings can help people to reconsider their words when they are given the opportunity to change," added the subsidiary firm of

A preview of the warning Instagram will issue after detecting an offensive textA preview of the warning issued by Instagram after detecting an offensive text

With photos, videos, Stories and long-lasting productions on IGTV, Instagram was under pressure to solve the problem of harassment after a series of notorious cases, including the suicide of the British teenager
Molly Russell

Adam Mosseri, the company's executive director, said of Instagram
"could do more" with respect to the problem, after announcing in June the system of automatic detection of offensive content.

The artificial intelligence system, which automatically detects a text considered potentially offensive, is based on previous reports made by users. This function will begin to be available gradually in some countries, and then be fully implemented worldwide in the coming months.