Meta and YouTube are using artificial intelligence (AI) to remove disturbing content from their websites, the BBC reported Thursday.
Meta and YouTube are using artificial intelligence (AI) to remove disturbing content from their websites, the BBC reported Thursday.
The report also said social media giants like Meta and YouTube, the world’s largest video platform, are using AI to remove “disturbing” content from their websites. But experts are skeptical of the move. They believe AI technology can’t distinguish between videos of violence and videos that could be evidence of human rights violations.
The report also said social media giants like Meta and YouTube, the world’s largest video platform, are using AI to remove “disturbing” content from their websites. But experts are skeptical of the move. They believe AI technology can’t distinguish between videos of violence and videos that could be evidence of human rights violations.
subscribe and read more
For example, former journalist Ihor Zakharenko tried to post videos of human rights abuses by Russian troops in Ukraine on Facebook and Instagram. The video was quickly deleted. The BBC reported that Zakharenko tried to upload the videos again using a dummy account, but three of the four videos were taken down in just 10 minutes.
Content moderation policy
The social media company’s community guidelines allow the removal of graphic violence videos, although such videos may also be allowed if they are shared in connection with important and newsworthy events. also said. But experts say AI technology cannot make such subtle decisions.
“We don’t allow graphic violence, and we may remove intensely violent videos and images to help keep Instagram appropriate for everyone. Sharing in connection with important and newsworthy events and may be allowed if this image is shared for the purposes of denunciation, awareness, or education,” according to Instagram’s community guidelines.
“Violent or gory content intended to shock or disgust viewers, or content that encourages others to commit violent acts, is not allowed on YouTube.”
YouTube’s guidelines do not allow traffic accidents, natural disasters, war aftermath, terrorist attack aftermath, street fighting, physical attacks, self-immolation, torture, dead bodies, protests or riots, robbery, medical procedures, or any other impact or impact I’m referring to a specific scenario, such as a similar scenario with disgusting viewers.
Tightrope walking
Meta and YouTube are committed to testifying and balancing their respective obligations to protect users from harmful content. But in reality, using common artificial intelligence tools to remove all kinds of violent content is far from maintaining a balance.
Experts cited by the BBC say social media companies are either eliminating AI from their content moderation process and relying on human judges to decide what content to remove, or relying on videos of violence and potential violence. He suggests that we should either develop an AI that is better at distinguishing between certain videos. Evidence of human rights violations.
