The social media app Yubo has garnered a following among younger users, many of whom are members of Gen Z. The app has installed multiple safety features to help ensure that these users are interacting in safe and responsible ways. The most impressive of these safety features are two that leverage artificial intelligence.
Age Verification With Artificial Intelligence
Artificial intelligence has been built into the onboarding process, where it’s used to verify the ages that users provide when signing up.
Users generally must be at least 13 years old to use social media, and some platforms (including Yubo) classify user accounts into separate adult and minor categories. The obvious issue with this method is that users can lie about their age when signing up.
To prevent users from lying about their ages, Yubo has partnered with Yoti to create an AI age verification algorithm. The algorithm relies upon an uploaded photograph and video to confirm users’ imputed ages. The age verification is based on the still photo, and the photo is verified by checking the video.
Using AI to verify age has proven to be effective. The algorithm estimates the age range with nearly 99 percent accuracy, and it often gets ages within about 1.5 years. Moreover, these numbers hold true regardless of gender, skin tone, or age range.
Turning to artificial intelligence for age verification was a new idea when Yubo first started using it in 2019. Now, other platforms are following suit. Instagram plans on AI-verifying all user ages by December 2022, much like Yubo does.
Live Content Monitoring With Artificial Intelligence
Yubo also uses artificial intelligence to monitor content as it’s streamed to the platform. Most of the platform’s content consists of live streams that small groups connect around, so real-time content moderation is especially important.
Artificial intelligence is constantly monitoring what people stream on the platform. The software takes screenshots of streamed content each second and quickly analyzes the screenshots for potentially inappropriate content. Inappropriate content can include anything with nudity, guns, violence, blood, or drugs.
People Can Still Intervene
While utilizing artificial intelligence in helpful ways, Yubo hasn’t turned over all safety processes to computer programs. Both age verification and content moderation allow people to intervene directly.
The small ~1 percent of users who don’t have their age automatically verified by the AI software can still prove their age. A secondary option to provide additional verifying information ensures no one who’s of appropriate age is prevented from using the platform.
Any content that’s flagged by AI software is forwarded to a team of safety specialists. These are trained individuals who review the content and take any action that’s appropriate. Specialists are working all day and every day, so that content is continually being monitored at all times.
Using AI and People
By combining AI software with real-person intervention, Yubo is creating a safe platform for people to share live streams on.