What Misinformation Looks Like During an Emergency

# What Misinformation Looks Like During an Emergency

A flood warning is issued for one district. Within an hour, a voice note circulates claiming the dam has already failed — it has not. A screenshot of an evacuation map from a previous year's cyclone is shared as current. Someone in a fluorescent vest knocks on doors telling residents to leave immediately; they represent no official agency. A relative forwards a message recommending a household substance for respiratory symptoms — a recommendation no health authority has made and that poison control centers have warned against.

None of these are unusual. They are recurring features of the information environment during emergencies.

This article is not medical guidance. It does not tell readers what to do in an emergency. It examines what research and authoritative communication frameworks tell us about the forms misinformation takes during crises, why it spreads faster than at other times, what documented harm it has caused, and what that implies for the professionals who must communicate through — and against — a contaminated information environment.

The framing matters. Misinformation during emergencies is often discussed as a problem of gullible individuals sharing bad links. The evidence suggests a different diagnosis: misinformation is a systems failure produced by predictable interactions between information voids, acute stress, social sharing dynamics, and institutional communication gaps. Understanding the pattern is prerequisite to reducing harm.

A taxonomy of emergency misinformation

Misinformation during crises is not one phenomenon. It is a family of related failures that share structural features but differ in origin, intent, and correctability. Research across public health infodemic management, crisis informatics, and emergency risk communication identifies several recurring categories.

False or impersonated official instructions

Among the most directly dangerous forms are fabricated evacuation orders, false "all clear" messages, and impersonation of firefighters, police, or government workers. Documented U.S. cases include door-to-door false evacuation orders during Hurricane Harvey (2017), individuals in fluorescent vests falsely clearing evacuation zones before Hurricane Florence (2018), and a man impersonating a firefighter ordering evacuations during an Arizona wildfire (2024) — causing panic in neighborhoods where no official evacuation had been ordered. Similar patterns appeared during Oregon wildfires (2013), where sheriff's officials reported false door-to-door evacuation claims potentially intended to facilitate theft after residents left.

These cases share a feature: they exploit the authority gradient of emergencies. When people are frightened and official channels are overloaded, a confident person in uniform — or a message claiming to be from an official account — can override verification behavior that might prevail under normal conditions.

Hypothesis (not independently validated at scale): Impersonation incidents are undercounted because they are handled locally as fraud or disorderly conduct rather than aggregated as emergency communication failures.

False remedies and hazardous health claims

During health emergencies, misinformation frequently takes the form of unverified treatments promoted as cures or preventatives. The COVID-19 pandemic produced extensively documented examples: promotion of chlorine dioxide — an industrial bleach — as a "miracle cure" in Bolivia and other countries, despite warnings from the Pan American Health Organization and documented poisonings. Associated Press reporting from Cochabamba described long queues for small bottles of the toxic substance even after Bolivia's health ministry reported poisonings. A geospatial analysis in the Greater Boston area found that social media discussion of untested cleaning-product remedies correlated with increased poison control center calls about household cleaners two to four days later — suggesting a measurable link between online misinformation and harmful behavior, though correlation is not causation.

These cases illustrate a pattern: when official treatment options are limited, frightening, or poorly communicated, alternative claims offering certainty and agency spread rapidly — especially when promoted by trusted community figures or packaged as suppressed knowledge.

This article deliberately avoids describing preparation methods or dosages for any false remedy. The safety principle for communicators is description of harm and pattern, not replication of instructions.

Outdated advice recirculated as current

Not all harmful emergency information is newly fabricated. Some is previously debunked or context-specific guidance that resurfaces during a new event — often via messaging app forwards. A first-aid technique deprecated years ago. A flood safety tip from a different geography. An earthquake myth that geologists have repeatedly corrected. The content may have been plausible in its original context; its harm comes from being detached from time, place, and current evidence.

This pattern is particularly common on WhatsApp and similar closed-group platforms, where forwards strip away original dates and sources. Crisis informatics literature has historically focused on public platforms like Twitter; a significant evidence gap exists for closed-group recirculation dynamics, though practitioner toolboxes now explicitly recommend monitoring WhatsApp and Telegram.

Misattributed media and fabricated footage

During disasters, images and videos from previous events — or from different locations — are widely shared as depicting the current crisis. AI-generated and manipulated media add a newer variant: fabricated disaster footage, synthetic voice alerts, and deepfaked statements attributed to officials. The JMIR Infodemiology narrative review (2025) notes that misidentified disaster footage (wrong location, wrong event, wrong date) remains a persistent category with consequences for situational awareness and resource allocation.

Rumors as sensemaking — not always malicious

Academic rumor research, particularly the work of DiFonzo and Bordia, distinguishes rumor from deliberate disinformation. Rumors are unverified information statements circulating in ambiguous or threatening situations, often serving collective sensemaking — people trying to understand what is happening when official information is incomplete. Some rumors later prove accurate. Many do not. The safety problem is identical in the moment: unverified information drives action before verification is possible.

Crisis informatics researchers including Kate Starbird and colleagues have documented how this "collective sensemaking" process plays out on social media during events such as the 2013 Boston Marathon bombings — with distinct propagation "signatures" for different rumors, and with both amplification and crowd-driven challenging occurring simultaneously.

The implication for emergency managers: not every false claim originates from a bad actor. Much of it originates from frightened people trying to help — which changes the appropriate response from condemnation to verified replacement.

Algorithmic amplification and information overload

During crises, platform algorithms, news coverage, and the sheer volume of content can amplify emotionally charged false claims faster than sober corrections. The WHO defines an "infodemic" as the overabundance of information — including mis- and disinformation — during a health emergency. The concept applies beyond health: any emergency can produce more information than any individual or institution can evaluate, creating conditions where speed beats accuracy in sharing behavior.

Why misinformation spreads faster during crises

Three mechanisms — institutional, psychological, and technological — interact during emergencies in ways that rarely coincide during normal times.

Information voids

The World Health Organization's emergency communication guidance states plainly that late announcements and information voids allow rumours to fill the gap. CDC's Crisis and Emergency Risk Communication (CERC) framework lists among communication failures that worsen public outcomes: mixed and conflicting messages from multiple sources, late release of critical information, and unaddressed or uncorrected myths and rumors.

When people need to decide whether to evacuate, shelter, seek care, or help a neighbor, absence of timely official information does not produce silence. It produces substitution. Unverified community information becomes the best available signal — a dynamic Oh, Kwon, and Rao documented in their analysis of Twitter during the Mumbai terrorist attacks, Haiti earthquake response, and other social crises, where information with no clear source was the strongest predictor of rumor activity.

Altered cognition under stress

Related research on emergency information design (see the platform's article on designing guidance for people under stress) establishes that acute stress degrades comprehension, narrows attention, and impairs working memory. CDC's CERC psychology guidance adds crisis-specific information-processing patterns: people simplify messages, hold onto existing beliefs, seek confirmation before acting, and often believe the first message they receive — even if more accurate information follows.

This has direct misinformation implications. A frightening false claim received first may persist even after correction — a phenomenon well documented in misinformation correction literature. Van der Meer and Jin's experimental work on crisis misinformation found that while both simple rebuttals and detailed factual elaborations can reduce false beliefs, providing a strong alternative narrative — not merely negating the false claim — is important for both belief correction and mobilizing protective action.

Hypothesis: The interaction between stress-induced first-message stickiness and rapid social sharing creates a narrow correction window — likely measured in hours, not days — though precise timing thresholds vary by event type and are not yet established by controlled field studies.

Social sharing dynamics

Crisis informatics research consistently finds that social media elevates the public into an information "first responder" role — faster than institutions, but less reliably. Personal involvement and anxiety drive sharing. Closed groups create rumor cascades where content is stripped of provenance. Starbird's research on official account participation found that timely denials from credible institutional accounts can slow rumor spread — but mainstream media and "breaking news" accounts also participate in rumoring, not only amateur users.

The speed asymmetry is structural: sharing requires one tap; verification requires multiple steps — checking source, cross-referencing official channels, assessing date and context. During a crisis, the cost of verification (time, cognitive effort, connectivity) rises just as the cost of sharing (social bonding, anxiety reduction, perceived helpfulness) falls.

Documented consequences — without amplifying the claims

The harm from emergency misinformation is not theoretical.

Physical harm from false remedies: Documented poisonings and at least one death linked to chlorine dioxide promotion during COVID-19 in South America; poison control call increases correlated with social media remedy discussion in Boston. These are among the clearest quantitative links between misinformation and health harm.

Security and displacement harm from false evacuations: Documented impersonation incidents causing unnecessary panic, displacement, and — in cases authorities suspected — facilitation of looting. Even when no physical injury occurs, false evacuation imposes real costs: traffic congestion on evacuation routes, shelter overcrowding, erosion of trust in legitimate future orders.

Resource misallocation: CDC's CERC introduction notes that crisis-related psychological dynamics, left unmitigated by trusted communication, can produce misallocation of treatments based on demand rather than need, unreasonable avoidance behaviors, and propagation of damaging rumors directed at people or products.

Trust erosion: WHO and CDC both identify mistrust as a force multiplier for misinformation. If people affected by a crisis do not trust official sources, they may accept incorrect information from sources they perceive as more credible — including community leaders, celebrities, or anonymous forwards. Mistrust is both a cause and a consequence of misinformation failures, creating a compounding cycle across sequential events.

Communicators should describe these patterns and their consequences without repeating the specific false claims in ways that re-expose audiences to them — a discipline WHO's 2024 operational toolkit implicitly supports through its emphasis on replacement messaging and outreach rather than mere refutation.

Who is vulnerable — and why that is the wrong question

It is tempting to ask "who falls for misinformation?" and implicitly blame insufficient education or critical thinking. The research does not support that framing.

Vulnerability during emergencies is situational, not dispositional. Anyone experiencing high anxiety, incomplete information, time pressure, and social pressure to act or share is operating in conditions that predictably degrade verification behavior — including trained professionals, as information overload research on emergency managers suggests.

Specific structural vulnerabilities include:

• Limited connectivity or digital access, which creates dependence on second-hand forwards and verbal chains
• Low health or disaster literacy, including difficulty evaluating source credibility or recognizing outdated content
• Language barriers, which push people toward community-language channels that may lack official verification
• Prior institutional mistrust, rooted in historical experience rather than individual skepticism
• Social network position, where close-knit groups amplify internally trusted but externally unverified claims

The JMIR Infodemiology review (2025) notes that older adults may unknowingly spread misinformation more frequently — a finding that supports targeted education rather than blame, since sharing often reflects generational channel habits and trust in peer-forwarded content.

The design implication aligns with emergency information research: systems should assume the least prepared verifier, not the ideal critical thinker. Verification cannot be delegated entirely to individual cognition during acute crisis.

What emergency responders and communicators can do

WHO's infodemic management framework organizes institutional capacity into four domains: prepare and monitor, detect, intervene, and strengthen. Its 2024 operational toolkit specifies a five-step process for false information: signal detection, verification, risk assessment, response design, and outreach. CDC's CERC guidance on rumors recommends monitoring traditional and social media, correcting damaging misinformation immediately through appropriate channels, and regularly updating information outlets to prevent speculation.

Synthesizing across sources, the evidence-supported strategies include:

Fill information voids before rumors do

Speed and accuracy are not opposites in planning — they are trade-offs managed through pre-positioned message templates, pre-identified spokespeople, and pre-mapped trusted community channels. WHO's outbreak communication planning guide (still widely referenced) emphasizes trust, early announcement, and transparency as rumor prevention — not merely rumor response.

Monitor to detect — not to surveil communities punitively

Infodemic surveillance, as described in CDC's Emerging Infectious Diseases analysis, can function as early warning for public health and communication response — identifying trending false claims before they peak. Ethical implementation requires transparency, privacy safeguards, and use for correction rather than punishment. The AI & Society disaster risk management toolbox (2025) recommends real-time monitoring across public social platforms and acknowledges the monitoring challenge of encrypted messaging.

Correct with narrative, not just negation

Research on crisis misinformation correction suggests that detailed factual elaboration — providing an alternative explanation that helps people abandon the initial false narrative — often outperforms brief denial alone, particularly for mobilizing protective action. This tension with CERC's emphasis on simplicity under stress is real. Resolution hypothesis: use minimal messages for immediate life-safety actions ("evacuate zone X"); use elaborated corrections for persistent false beliefs that continue after immediate danger passes. This phased approach is consistent with both bodies of guidance but requires empirical validation in field settings.

Deploy trusted messengers

CDC's guidance on addressing vaccine misinformation — transferable in structure if not in content — emphasizes that some populations will not trust public health institutions but will trust religious leaders, community organizations, or local figures. Correction delivered through pre-existing trusted relationships outperforms identical content from unfamiliar institutional accounts — a finding consistent across public health and crisis communication literature.

Official denial when credible — and fast

Starbird's research on Boston Marathon rumors suggests official account denials can slow spread — but timing and credibility prerequisites apply. An organization that has not communicated early may lack the standing to correct late.

Digital literacy as preparedness infrastructure — not a side project

The disaster literacy framework developed after the 2023 Turkey earthquake argues that digital and media literacy should be a central competence in disaster preparedness — including skills for verifying sources, evaluating media content, and recognizing disinformation. This is not a call to teach civilians to become fact-checkers during a flood. It is a case for building verification habits, trusted source directories, and community information networks before crises occur.

Practitioner accounts from U.S. cooperative extension digital skills programs describe disaster preparedness training that includes identifying trusted sources (FEMA, Red Cross, local emergency management), using smartphone emergency features, and recognizing common forward patterns — integrated into general digital literacy rather than standalone media studies.

The JMIR Infodemiology review recommends public education campaigns on digital literacy and proactive debunking as components of disaster preparedness plans — alongside technological detection — noting that misinformation spreads more rapidly when official communications are unclear or delayed. Literacy and institutional clarity are complementary, not substitutes.

What we do not yet know

Honesty requires naming the gaps.

There is limited causal evidence linking most categories of emergency misinformation to quantified mortality at population scale — except in relatively clear cases like toxic remedy ingestion and poison control correlations. The full burden of false evacuation, misallocated sheltering, and delayed appropriate care is likely substantial but methodologically difficult to measure.

Crisis informatics research remains skewed toward public, English-language, Twitter-era data. WhatsApp-forward ecosystems — critically important in India and much of the Global South — are understudied relative to their harm potential.

AI-generated emergency misinformation is evolving faster than peer-reviewed evaluation. Hypothesis: Synthetic alerts and fabricated official statements will become a standard category within 2–3 years; institutional verification infrastructure is not yet uniformly prepared.

Retroactive correction — reaching people who already shared false information — lacks robust evidence for scalable effective methods. Platform "recall" mechanisms are weak for forwarded content.

The tension between rumor-as-sensemaking and rumor-as-harm remains unresolved in operational doctrine. Over-aggressive debunking of community-generated early reports could suppress useful intelligence; under-aggressive correction allows harm. Professional judgment and local context dominate where evidence is thin.

The question beneath the question

Emergency management invests in physical infrastructure — levees, shelters, dispatch systems, stockpiles. It invests less systematically in information infrastructure: pre-crisis trust, monitoring capacity, correction protocols, verification literacy, and channel relationships with communities that will not read press releases.

Misinformation during emergencies is not an aberration. It is the default output of a system where demand for certainty exceeds supply of verified information, where stress degrades the cognitive tools people use to evaluate claims, and where sharing is faster than checking.

The question this article raises is not whether misinformation will appear during the next emergency. It will. The question is whether the institutions responsible for public safety treat the information environment as part of the emergency itself — with the same seriousness as water levels, fire spread, or caseload curves.

If they do not, the failure will look like public gullibility. It will look like irresponsible sharing and viral stupidity. It will rarely be recognized for what the evidence suggests it is: a predictable systems outcome in an environment where false information is often the most available information — and where correcting it requires preparation that cannot be improvised after the dam breaks, the fire crosses the ridge, or the first forward goes viral.