Zicore has unveiled a solution that enables real-time management of large-scale illegal content faced by companies.

On the 5th, Gcore launched the 'Gcore AI Content Moderation' solution that helps detect and manage audio, text, and user-generated video content in real time.

The solution will enable online service providers to automate sanctions against inappropriate content without having to directly apply AI/ML. It will also “improve user safety and comply with country-specific regulations such as the EU’s Digital Services Act and the UK’s Online Safety Act,” Zicore explained.

All online platforms that host content accessible to children are required to have measures in place to moderate user-generated content to protect viewers from violent and illegal content.

Zicore's 'AI Content Moderation Solution' filters out harmful content from a large amount of data.

First, it automates the review of video content streams to flag inappropriate content. If necessary, alert a human moderator. Review begins within seconds of a video or live stream being posted.

Zicore's 'AI Content Moderation Solution' is equipped with the following functions: △Using cutting-edge computer vision models to detect, segment, and classify objects; △Optical Character Recognition (OCR) to convert text displayed in videos into machine-readable formats; △Speech recognition algorithms to detect profanity; and △Aggregation of multi-model outputs.

Additionally, the ZigCor AI content moderation solution can be integrated into a company’s existing infrastructure via API. ZigCor added that it runs on ZigCor’s global edge network of more than 180 nodes with a total network capacity of more than 200Tbps, ensuring low latency, speed, and resilience for customers.

“It’s nearly impossible to staff and control costs to moderate massive amounts of content,” said Alexey Petrovskikh, Zcor’s head of video streaming. “Our new AI Content Moderation solution empowers enterprises to discover and manage content at a global scale, while also enabling humans to review flagged content. This is essential to helping enterprises comply with regulations while protecting their users, communities, and corporate reputation.”