Constant emotional alarm on social media is the natural low-energy state of the platform's engagement code.
Digital environments are built to favor high-arousal states like fear and anger as their default setting. Most people believe that online polarization is the work of specific bad actors or political conspiracies. This research shows that engagement-based algorithms drift toward alarm because it is the most efficient way to keep users active. Anxious users stay on the platform longer, creating a structural incentive for the software to keep them upset. This means the anxiety we feel online is a feature of the architecture, not just a result of the content.
Alarm as baseline
SSRN · 6637638
The dominant narrative on disinformation and political manipulation in digital environments presupposes, explicitly or implicitly, the existence of an intentional actor: someone who designed the state of social alarm for that purpose. This presupposition directs regulation toward specific actors while leaving the underlying infrastructure intact when those actors are dismantled. Sustained emotional activation, however, persists in the absence of documented disinformation operations, suggesting t