Engineering Epistemic Resilience in Information Ecosystems

Open your phone during a crisis, and the feed feels less like a reliable source of information and more like a bout of turbulent weather. Headlines contradict each other, rumors race ahead of official guidance, and each scroll chips away at your sense of what to trust. Health agencies, election bodies, and platforms have established fact-checking units, moderation teams, and rapid response plans; yet, every major outbreak or election still brings its own surge of rumors and doubt. The World Health Organization (WHO) now treats health-related disinformation as a direct threat to public health and safety.¹ During large outbreaks, falsehoods routinely travel further and faster than corrections, creating what WHO calls an “infodemic” that undercuts guidance and fuels fatigue and frustration.²

Economic models of online sharing suggest these patterns follow from the rules of the system rather than from isolated bad actors. In attention markets where people gain social rewards from engagement and platforms monetize time spent on site, there is a steady incentive to share content that evokes strong emotions and signals identity, even when these practices increase the risk of spreading misinformation.³ Empirical work on social media during the pandemic shows how this plays out: emotionally charged, low-quality health content leveraged these dynamics to reach huge audiences, while careful expert explanations struggled to keep up.⁴

Most debates still focus on content-level fixes. Remove the most harmful posts, label dubious claims, promote fact-checks, and add prompts that nudge users to slow down. These tools help with individual decisions. They do not fully address how the whole system behaves under sustained stress. When media infrastructures are thin, recommendation systems reward outrage, and trust in institutions is low, even well-designed nudges operate in a headwind. The deeper question is how to build information ecosystems that stay anchored to reality during shocks, adapt when evidence changes, and regain credibility after mistakes without centralizing control over what people see or say.