Last mentioned: Jul 6, 2026
The royal commission testimony underscores persistent brand-safety risks for advertisers on Meta and live-streaming platforms like Kick, where hate speech detection is inconsistently defined.
Cybersecurity professionals view Kick's reliance on outsourced moderators and subjective hate speech policies as a systemic vulnerability exploitable by threat actors to spread harmful content undetected.
The 'art vs. science' problem described by Kick highlights the limits of machine learning models in detecting nuanced hate speech in live video, as Meta's tiered policies also expose AI blind spots.