
Platform Algorithms Structurally Amplify Outrage Regardless of Safety Policies
Social media recommendation systems remain predisposed to amplify high-arousal content—outrage, fear, tribal solidarity—independent of stated platform policies. The 2024 UK riots exposed this gap: engagement-optimized design amplified inflammatory material regardless of terms of service. Closing the divide between policy and system architecture represents the substantive engineering challenge the Online Safety Act cannot yet address through regulatory oversight alone.
Published