YouTube has faced heavy criticism over what some believe is an inadequate level of protection for young users. Though the company has already taken a number of steps to shield young eyes from inappropriate videos, including launching a special app for kids, it plans to take things a step farther by using its artificial intelligence technology to automatically age-restrict videos that are detected as inappropriate for users under the age of 18. Embedding the content in third-party websites won’t get around the restriction, either.
YouTube detailed its plan to expand the use of its machine learning technology in a blog post today, explaining that it will soon use the AI to detect and automatically age-restrict videos that likely aren’t appropriate for young viewers. Ideally, YouTube creators will manually age-restrict their content when uploading to the service, but as YouTube notes, not all users take the time to do that.
To make up for the times when users fail to restrict their videos, YouTube has leveraged machine learning to find the videos and bump them to its human team for review. The newly announced change is a step up from that, removing the human review from the equation to instead automatically age restrict the video. Once restricted, viewers will need to sign in to their account to verify that they are old enough to watch the content.
Any age-restricted video that is embedded in a third-party website will require the user to click through and watch it on the YouTube platform. This will ensure that young viewers aren’t able to get around the age restriction by watching the content elsewhere, though this may be a bit annoying for everyone else.
Creators who disagree with an automatically age-restricted video change can appeal the decision to YouTube, which will bump the appeal to its Trust & Safety team to make a final determination about the video’s classification. YouTube attempts to reassure monetized creators by saying that it expects the new change will have ‘little to no impact on revenue.’