Harmful misinformation as a quantifiable risk in health content: WHO-aligned risk assessment framework for flagging potential harms and informing policy safeguards

Health misinformation undermines public decision-making, yet prevailing responses lean heavily on binary fact-checking that misses gradations of misleadingness and, crucially, the potential of downstream harms. This article presents the Misinformation Risk Assessment Model (MisRAM), a WHO-aligned framework that conceptualizes misinformation as a quantifiable risk in content and mediating systems, rather than a present/absent defect. Drawing on health-risk research traditions, MisRAM guides the identification and treatment of misinformation risk factors as informational hazards, whose risk contribution varies with intrinsic severity, cumulative load, exposure conditions, and recipient susceptibility. It maps four evaluative dimensions: inaccuracy, incompleteness, deceptiveness, and harm potential, with the latter requiring the highest protection and prioritized as an independent escalation trigger. This facilitates stratification into very-low through to very-high risk of misinforming recipients in ways that may lead to harmful outcomes, and is designed for both domain-led safety evaluation benchmarking and high-throughput automated detection. To inform risk thresholds and calibrated mitigation, the Severity of Potential Harmful Outcomes from Misinformation (SoPHO-M) Scale is introduced as a five-level taxonomy for classifying plausible harm severity (from negligible to fatal). The framework provides a basis for policy-relevant systemic risk assessments, informing governance protocols across digital services and AI systems, as well as regulatory compliance. Tiered safeguards are proposed for different misinformation risk levels in proportion to the likelihood and severity of their effects, ranging from positive incentivization and visibility support of low-risk, well-contextualized content, through friction and conditional visibility for moderate-risk content, to red-line protective controls at very-high risk where life-threatening harms are plausible. By moving beyond truth judgements towards harm-oriented risk estimation, MisRAM provides a roadmap for assessment-tool development and proportionate safeguard operationalization, with potential for extension to other high-stakes domains. The framework offers guidance to decision-makers on preventing serious harm without disproportionately burdening freedom of expression or information access across mediated systems.