Report shows that Jihadi content is still present on Facebook and Google
As we reported earlier, Facebook has claimed that their Artificial Intelligence (AI) tools are now detecting inappropriate jihadi content with great success. 99.5% of this content gets instantly removed by these tools. This is a big step forward since this number was at 96% previously.
All of these online jihadi content claims sound great in theory but, a recent report from the internet safety nonprofit Digital Citizens Alliance suggests that these AI tools still need a lot of improvement when detecting and removing terror content online.
This report explained how easy it is to bypass the AI security on Facebook and post images of dead bodies, calls to jihadi and ISIS members as well as Al Qaeda imagery on both Facebook and Instagram. This report came with a lot of screenshots of beheadings, terrorist recruitment, and images that are linked to accounts that are still active on these social media websites.
“It seems based on everything we know the platforms are stuck in a loop. There’s criticism, promises to fix, and it doesn’t go away,” says Tom Galvin, executive director of the Digital Citizens Alliance.
Working with researchers at the Global Intellectual Property Enforcement Center, or GIPEC, the Digital Citizens Alliance has uncovered a lot of evidence suggesting that terrorist content still thrives on these social media platforms. In this research they used artificial intelligence combined with human vetting to search for suspicious keywords and hashtags. From there, they found a lot of other accounts that are associated with the posts.
The research confirmed that a lot of images showing ISIS supporters can still be found on Instagram and Facebook. WIRED also investigated one account on Instagram. They found an image showing two men being beheaded by soldiers in black face masks. A couple of days later the image was removed but the account that contained a lot of photos with similar content was kept intact so it is not clear if Instagram removed the image. The most disturbing photos contained captions and hashtags written in Arabic including #Dads, #Girls, and #Cooking.
Facebook doesn’t have as many photos but textual posts can be found more often than you think. One particular post promotes violence and contains words like “kill the unbelievers,” accompanied by tips on how to do just that. WIRED did talk to a Facebook spokesperson that stated: “There is no place for terrorists or content that promotes terrorism on Facebook or Instagram, and we remove it as soon as we become aware of it. We take this seriously and are committed to making the environment of our platforms safe. We know we can do more, and we’ve been making major investments to add more technology and human expertise, as well as deepen partnerships to combat this global issue.”
Eric Feinberg, GIPEC’s founder, explains how Facebook isn’t doing enough to prevent posts like this. He stated that there are cases where Facebook removes a post or restricts an account but they leave it there. They should also investigate accounts linked to that one, the people who reacted to the post and participated in the comments as well. This is a service our latest intelligence platform, INVISO, offers.