Instagram will block suicidal content in the EU, but claims privacy laws limit scope

Almost two years ago we told how from Instagram they recognized that it was necessary to regulate the posts of self-harm after the suicide of a teenager. Since then it has rained a lot on the social network and on Facebook, and the company has been actively searching for this content with machine learning for a while, and then send it to human reviewers who make the final decision on the content.

Adam Mosseri, CEO of Instagram, tells in a publication that between April June of this year, 90% (almost 94%) of the self-harm content on which they acted was detected by technology without anyone reporting. They equal about 275,000 pieces of content. However, while highlighting these milestones, they affirm that due to regulation, until now they have not been able to use these techniques in the European Union, something that is now partially changing.

Instagram and Facebook can already automatically detect content with suicidal tendencies in the EU, but with limitations

What has changed in November 2020 in the community territory is that now Facebook and Instagram can find more content on suicides and injuries and make it less visible. Mosseri states that “we can now search for posts that are likely to break our rules on suicide and self-harm and make them less visible to the public. automatically remove them from places like Explore“.

In other words, what works has to be executed automatically, and in order for it to be eliminated, the technology they use must be very secure. The problem is that, with current regulation, the automated system is not capable of transmitting the contents to human reviewers, a fundamental fact so that more letters can be taken with the matter, follow-up or contact the authorities.

Tara Hopkins, head of Instagram policy for Europe and the Middle East, said “a compromise proposal is now being discussed with the Irish Data Protection Commission (DPC)”. From the DPC, they have stated that “Facebook has recently approached the DPC proposing a much more limited use of this tool for the elimination of content that contravenes ”, Y claim to be reviewing the regulation to give an answer.