Moral panic: Facebook is accused of insufficiently effective censorship

It took only a few months after the scandal with the "fake news" that spread through Facebook and allegedly helped in the election of a new US populist president. Facebook and Google News were blamed for distributing this news. Like, they did not check the sources of news and displayed the most popular headlines on the top lines in the tape. Facebook had to urgently make changes to the algorithm for forming the news feeds. It seems that this algorithm is so complex that no one clearly understands what news he will choose (individually for each reader).

Now Facebook is again accused of “criminal negligence”. This time the reason was the behavior of an American citizen who decided for some reason to kill a random person on the street. He took off on the phone, both fits and shoots him, and then uploaded the video to Facebook. There it was 2 hours 7 minutes was available for viewing by everyone. One of the subsequent posts with a copy of the video scored more than 1.6 million views . Facebook has undergone notable obstruction , and journalists are raising the issue of the responsibility of the world's largest social network for what information it helps to disseminate. The question is how cruel and effective can censorship be?

The crime was committed by 37-year-old Steve Stephens from Cleveland. Initially, the police reported that the man broadcast live on Facebook live, but later clarified that the recording was made previously on the same day, and only then uploaded to the live streaming site Facebook Live. However, this does not change the essence of the claims. Commentators in the media and social networks immediately slipped into a moral panic and began to ask the question: what responsibility is Facebook ?
')
Commentators point out that the problem of unauthorized transmission of extreme violence live appeared before. For example, in 1974, American TV presenter Kristin Chubbak committed suicide live during a television broadcast on ABC. On July 15, 1974, she kept a criminal chronicle, but in the eighth minute of the program, an overlay happened: the announced shooting story in the restaurant could not be shown for technical reasons. Then Kristin said: “By supporting Channel 40’s policy of displaying the freshest blood and intestines in full color live, you will be the first television viewers to be shown a live suicide attempt.” Having said that, the TV presenter took out a revolver, put it to her head and pulled the trigger. According to her transfer scenario, she had to commit suicide at the end of the program, but because of the technical overhead, the moderator decided to outrun the events.

Of course, the channel has never repeated this recording - and not so many viewers could see it live. Certainly less than the 1.6 million people on the social network.

Facebook management responded as quickly as it could. The video was deleted, and then the company published an official statement: “This is a terrible crime, and we do not allow this type of content,” the statement said. “We are working hard to maintain a secure environment on Facebook and keep in touch with law enforcement agencies in case of emergency when there are direct threats to physical security.”

The question arises: what else can you expect from Facebook? Pre-moderation of all published videos? Implementing a neural network with the recognition of shooting and the image of the gun in the frame? So far it is unrealistic to roll out in production, such technologies are tested at the level of scientific experiments.

The company cannot magically prohibit people from publishing videos or prohibit committing crimes. Although something can be done, of course. For example, to automatically block attempts to re-download this video by other users after the original was deleted. It is really quite real.

At the moment, Facebook, like many other sites, rely on crowdsourcing - that someone will mark the video as inappropriate to the rules of the site. He will go to the moderators for review - and they will block him. But this system by default assumes that someone will have to look at this video before it is blocked. In this case, Facebook received a message about the video with the murder in 1 hour and 45 minutes after the publication, and after another 23 minutes the account was blocked.

It has been said many times that in the conditions of an open information society on the Internet, it is impossible to impede the free flow of information - this is contrary to the very nature of the World Wide Web. It's like trying to stop the spread of gas that has fallen into the atmosphere. Before the video was removed from Facebook, it was already sold in hundreds of copies in all corners of the Internet. Pasta can not be pushed back into the tube, and access to countless copies of these videos can not be blocked. They can not be removed from the Internet, they will remain there forever.

Even if Facebook and others introduce a fantastic AI system to automatically recognize and block video live, this censorship can always be circumvented: change the sound of a shot, mask the gun in the frame, etc. Nothing can be done about this, it only remains to be reconciled, that is, to fight in the old manner - to punish after the fact after crimes, and not to try to prevent them in a magical way.

Since Facebook launched the live video service a little less than a year ago, it has been used repeatedly to broadcast police shootouts, rape , torture, beatings and a large number of suicides . But it was a planned murder, it seems, was now broadcast for the first time.