In October, the 17th annual National Bullying Prevention Month took place. Launched by PACER’s National Bullying Prevention Center, this month encourages NGOs, companies, governments and individuals to raise awareness about online bullying.
With the pervasive use of the internet and social platforms today, anti-bullying awareness efforts are especially important in helping young people stay safe and protected when interacting online.
Many social networks are working to put measures in place to fight against this type of harmful behavior. They have developed an army of AI filters capable of tracking specific words, sentences, emojis and even their contexts to help prevent it. While still useful, these measures are only the first step in reducing online bullying as they focus exclusively on images and text, leaving audio an unsolved challenge. With the rise of livestreaming and voice-based interaction, more and more exchanges are done through audio and social platforms have struggled to keep pace with moderation in this area.
However, recently, a live social discovery platform for Gen Z called Yubo has launched what might be a solution to the issues surrounding audio moderation. It announced the deployment of its audio moderation technology – in an industry first – to four of its main English-speaking markets: the U.S., the UK, Canada and Australia
Yubo’s audio moderation system was developed in collaboration with Hive, a company specializing in AI solutions for moderating visual and audio content. This new system transcribes audio conversations from voice to text and the transcriptions are then analyzed to detect any harmful or inappropriate words or phrases.
This article from Make Use Of explains that there are still many false positives, as the technology is still very new. Over time, the algorithm will learn to distinguish between, for example, lyrics in a rap song and instances of harmful speech that is considered offensive. Regardless, this innovation is an effective start to tackling online harassment a little better.
And if, like us, you’ve been wondering why it’s Yubo that’s undertaking this enormous challenge and not a big social network like Facebook or Instagram, here is their reasoning – Yubo’s team is committed to being at the forefront of safety innovation and moderation. One of Yubo’s main features on the app is the livestreams, and with 99% of its users being Gen Z between 13 and 25 years old, they know that teenagers are particularly subject and vulnerable to online harassment. At Yubo, they are constantly developing and implementing new safety features to help keep their users as safe as possible, which is why they are leading the charge when it comes to a complicated issue like audio moderation.
Online safety is essential, and it can only be strengthened by taking direct action against the risks and dangers that come with socializing online. With Yubo’s latest audio moderation, privacy is also taken into account. In fact, no human will actually listen to or read the conversations transcribed. Yubo explains that only transcripts displaying harmful words or phrases that go against its Community Guidelines are flagged and reported to its team of Safety Specialists. Everyone deserves to be able to enjoy the internet in peace, regardless of someone’s gender, sexuality, background or ethnicity that may make them a target for cyber-bullying. Finally, what online harassers say will be held against them.
It will be interesting to follow the evolution of this technology – especially its deployment in other markets – and there are any major changes to online behaviors following this new development.