The four firms will create a common database containing “fingerprints” of digital content removed from their platforms.
Facebook, Microsoft, Twitter, and YouTube – a subsidiary of Google/Alphabet – were announced Monday, December 5, a partnership globally, aiming to identify more quickly the ” the content to a terrorist “ on their platforms, and thus hinder their diffusion. The four american companies agreed to create a common database that includes the ” fingerprints “ digital photos or videos of propaganda and recruitment removed from their platforms.
No message, however, will be removed or blocked automatically : it will be up to each enterprise to assess if the identified content violate its own rules. Each will also decide regardless of the content it decides to add to the common base. They specify want to start with images and videos, ” the most extreme and egregious “, and therefore ” the most likely to violate the rules of all enterprises “.
Read also : Incitement to hatred, nudity,… Facebook to clarify its policy of censorship of the contents
” Each platform is different ”
Microsoft, Facebook, Twitter and YouTube say they in addition want to consider how to involve other firms in the future, but insist on the fact that each of the partners will continue to be treated as independent requests for information or removal of content by governments or law enforcement.
Read also : Against the propaganda of jihadist online, the government turns to the private sector
The initiative comes as the United States, the european Commission and a number of other countries have stepped up calls in recent months that the social networks are intensifying their fight against the propaganda of the jihadist online.
After these calls, Twitter has notably announced to have been suspended since mid-2015 more than 360 000 accounts promoting terrorism, and has observed a decrease in the use of its platform by users supporting the jihadists.
Sinead McSweeney, vice-president in charge of public policies of the social network in the region Europe/Middle East/Africa, said in a message sent to Agence France-Presse that a large part of the suspended accounts had been detected by technical means, but that the sharing of digital fingerprints with the common database would be manually and ” periodic “.
” there is not a single approach suitable to all to tackle this type of material ; each platform is different. “
Read also : Why the moderation teams of Twitter are not more responsive
No algorithm ” magic “
most social networks prohibit in their rules of use the posts calling for violence or making the apology of terrorism.
Even if they have improved their tools for automatic detection, the actors in the sector to reaffirm regularly have no algorithm ” magic “ to identify content issues and rely heavily on the reporting by their users infringing material.
Read also : associations étrillent the moderation of Twitter, YouTube and Facebook
The online distribution of content, ” terrorist “ is ” a pressing problem that requires attention of technology companies “, and therefore justifies a joint initiative announced Monday, has risen to one of the participating firms. However, she stressed the fact that this did not represent a ” new standard “.
No comments:
Post a Comment