The TrustLab Blog

Moderating "how-to" content is a unique trust & safety challenge. Unlike traditional speech moderation, instructions often face the "dual-use problem"—the same info can be harmless or dangerous depending on context. From weapon-making instructions and self-harm methods to AI-generated risks, there’s no silver bullet. Success relies on analyzing patterns, context, and industry-wide collaboration.

The "Com" subculture has birthed a dangerous threat: the 764 network. Falling under Nihilistic Violent Extremism (NVE), this group has no political agenda, seeking only the absolute destruction of society. Infiltrating gaming and social platforms, predators blackmail vulnerable youth into violence. Traditional enforcement is failing; cross-platform intelligence is vital.

War reshapes Trust & Safety – blurring online/offline harm, exposing system gaps, and demanding faster, more human-focused, proactive responses.

What do elections, slang, and sports betting have in common? They all prove how local culture can make or break digital safety. In our new TrustLab blog, we explore real-world examples—from moderating Portuguese slurs to identifying hate speech masked by local humor—that show why AI alone isn’t enough.

Explore the evolution of content moderation, comparing manual and automated approaches. Learn about hybrid solutions balancing efficiency and nuance in online safety.

Discover expert insights on content policy development and enforcement. Learn strategies for simplifying moderation, tackling misinformation, and improving online safety.

Explore the future of content moderation through AI-human collaboration. Tom Siegel talks about how this approach improves accuracy, protects moderators, and boosts online safet

Explore key challenges facing Trust & Safety teams in 2024, including emerging threats, scaling issues, and the need for proactive approaches in online content moderation.

Explore the critical balance between content and actor-level signals in online safety. Learn how Trust & Safety teams can leverage integrated approaches to combat harmful content and bad actors effectively.