Roblox Doubles Down On Child Safety with Advanced AI Moderation

Roblox pushes for child safety with advanced AI moderation tools for text, voice, and visual content.
Roblox Time Out Ban
Image: Roblox Corporation

Roblox is improving its safety measures for kids’ chats by using AI. This decision comes as the platform adjusts to a growing audience, including older users. There are billions of chat messages sent each day and a wide range of user-created content. So, Roblox faces a real challenge in moderating the platform to guarantee a safe and friendly environment for players.

Roblox’s safety approach is centered on a combination of artificial intelligence and human knowledge. The company’s text filter, refined for Roblox-specific language, can now identify policy-violating content in all supported languages. This is particularly important considering the recent launch of real-time AI chat translations. Additionally, an original in-house system moderates voice communications in real-time, flagging inappropriate speech within seconds. To complement this, real-time notifications inform users when their speech approaches violate Roblox’s guidelines, reducing abuse reports by over 50%.

Roblox carefully moderates visual content using computer vision technology to evaluate user-created avatars, accessories, and 3D models. Items can be automatically approved, flagged for human review, or rejected when necessary. This process also examines the individual parts of a model to prevent potentially harmful combinations from being overlooked.

This is just one of the many things Roblox has done. They’ve supported a bill in California for child safety, which is a big deal. However, even with the heat on the pay terms, they don’t think they’re exploiting children with their pay methods.

The company is creating multimodal AI systems that are trained on various kinds of data to improve accuracy. This technology helps Roblox identify how different content elements could lead to harmful situations that a unimodal AI might not catch. Essentially, it is a bunch of AIs working together instead of a single AI expected to handle everything. For instance, an innocent-looking avatar could cause issues when paired with a harassing text message. A good example is when a Netflix logo is used as a profile image for a TikTok user, and then the six-letter username is a banned word that starts with N but doesn’t use an N (because the profile image does that for them).

While using automation is important for handling a large amount of moderation work, Roblox stresses the ongoing need for human involvement. Moderators adjust AI models, deal with complicated situations, and guarantee overall precision. Specifically for serious investigations, such as those related to extremism or child endangerment, specialized internal teams take action to identify and handle the bad people.

Jorge A. Aguilar

Jorge A. Aguilar

Jorge A. Aguilar, also known as Aggy, is the current Assigning Editor.

He started his career as an esports, influencer, and streaming writer for Sportskeeda. He then moved to GFinity Esports to cover streaming, games, guides, and news before moving to the Social team where he ended his time as the Lead of Social Content.

He also worked a writer and editor for both Pro Game Guides and Dot Esports, and as a writer for PC Invasion, Attack of the Fanboy, and Android Police. Aggy is the former Managing Editor and Operations Overseer of N4G Unlocked and a former Gaming editor for WePC.

Throughout his time in the industry, he's trained over 100 writers, written thousands of articles on multiple sites, written more reviews than he cares to count, and edited tens of thousands of articles. He has also written some games published by Tales, some books, and a comic sold to Telus International.

More Content

Comments

Leave a Comment

All comments go through a moderation process, and should be approved in a timely manner. To see why your comment might not have been approved, check out our Comment Rules page!

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.