X

YouTube Cracks Down On Pseudo-Kid-Friendly Videos

YouTube has been hit with a wave of surreal and inappropriate videos targeting kids lately, and though the videos have become an internet mystery of sorts, YouTube is taking decisive action against the videos and their creators that implies that the video giant is content to leave the mystery unsolved, as long as the videos are gone. The firm is taking a five-pronged approach meant to not only demonetize those videos and ban their creators, but also to involve authorities and experts to investigate any possible cases of child abuse or endangerment involved in the videos, and reining in the sometimes wildly inappropriate comments that can pop up on these sorts of videos and others featuring minors.

The first big step up in enforcement is Google working closely with international law enforcement agencies. Any videos that feature minors in live-action and seem to depict abuse, neglect, endangerment, or other such behaviors will be reported and investigated, even videos that may seem innocuous, or may seem like the uploader’s intentions were not malicious. This is a stepping up of enforcement on a policy already in place, and Google is pulling in artificial intelligence by using machine learning techniques to more readily identify and flag these videos. The company is also using the same algorithm to step up enforcement of its rule that such content be demonetized, and it has successfully demonetized or removed over 3 million such videos thus far, as well as an additional 500 million or so since getting AI involved.

As far as comments go, YouTube seems to be done with simply weeding out bad commenters at this point – the new policy dictates that a single inappropriate comment on a video with a minor will mean that comments on that video will be disabled, and the commenters that caused it will find themselves investigated. Finally, Google is partnering up with experts and doubling the number of Trusted Flaggers in charge of content involving or targeted to minors. Naturally, the work of these humans will help to train Google’s AI to do the same work. These videos pop up in massive numbers and very short order, and usually amass thousands or even millions of views before being caught, though some or most of those may be fraudulent or automated. In any case, a stepping up of enforcement in this area is as good a way as any to bring the saga to an end, and the involvement of international law enforcement may well result in the mystery actually being solved.