BOULDER, Colo. - ColoradoDesk -- Stream, an emerging leader in content moderation, today announced the expansion of its AI-powered moderation platform with real-time video and audio moderation for live calls, livestreams, and other live media experiences.

The new capability extends Stream's existing moderation suite, which already covers text, image, and custom content, into live video and audio, making Stream one of the few platforms that can moderate an entire real-time communication stack from a single dashboard.

"Moderation doesn't stop at chat," said Thierry Schellenbach, CEO and Co-Founder of Stream. "When your product includes live video and audio, you need moderation that works in real time, not after the fact. This closes the loop for teams building any kind of live experience."

How It Works

Stream's call moderation analyzes video keyframes and audio transcriptions in real time during active sessions. When violations are detected, the system responds with configurable, escalating actions, from warnings to video blurring to removing a user from the call, without requiring manual moderator intervention.

More on Colorado Desk

• USA Med Bed Helping Home Care Patients with Refurbished Hill Rom Hospital Beds
• Sobreseimiento de Nicolás dos Santos y Jorge Méndez expone demandas millonarias a Paraguay y boicot a la Hidrovía
• Governor Polis and Colorado Congressional Delegation Applaud New Mission Assignment for Colorado National Guard Members
• CAPHRA warns Southeast Asia not to repeat Australia's nicotine policy failure
• Milo3D.ai Launches Free AI 3D Model Generator That Turns Text and Images Into Game-Ready 3D Assets in Seconds

Key capabilities include:

• Real-time video keyframe analysis for detecting visual content violations such as nudity, violence, and QR code spam
• Live audio transcription analysis using harm detection engines
• Configurable escalation sequences that progress automatically (e.g., warn → mute → kick)
• Cross-session violation tracking that enables enforcement actions against repeat offenders across multiple calls
• A unified moderation dashboard where flagged calls appear alongside text and image review queues

Built for Trust & Safety at Scale

Call moderation is fully configurable through Stream's Rule Builder or via API, giving Trust & Safety teams granular control over detection thresholds, confidence scores, cooldown periods, and escalation logic. Webhook support enables integration with external T&S tooling and incident response workflows.

Availability

Real-time video and audio moderation is available now. To learn more, visit getstream.io/moderation.