Introduction

When a hurricane makes landfall or a wildfire spreads, the immediate focus is on physical damage. But in today’s environment, disasters don’t just play out in the real world — they also unfold online.

Our latest case study, Weaponized Weather, shows how harmful narratives surge in the hours and days after disasters, filling the information vacuum before official updates are available. Using insights from Logically Intelligence, we analyzed more than 76,000 online posts tied to U.S. natural disasters between February 2024 and July 2025.

Here’s what we found:

• A small group of accounts repeatedly drive the majority of narrative amplification.
• State-linked outlets often seed stories before domestic influencers pick them up.
• Familiar conspiracy themes — FEMA failure, weather manipulation, land grabs — reappear across multiple disasters.
• AI-generated content is accelerating the speed and scale of these narratives.

Why it matters:

• For enterprise comms and risk teams, these narratives can erode trust with employees, customers, local communities and investors and create operational threats — especially when operations are tied to affected regions.
• For government agencies, the same narratives complicate crisis response, undermine institutional credibility, and increase pressure on frontline teams.

Understanding how these narratives take shape — and how quickly they spread — is critical for anyone responsible for protecting public trust, organizational resilience, or community safety.

Report

In the wake of recent U.S. natural disasters — from hurricanes and wildfires to floods and tornado outbreaks — online narratives have surfaced with striking consistency, coordination, and speed.

This report draws on more than 76,000 online posts from February 2024 to July 2025 and leverages Logically Intelligence (LI)— our proprietary threat detection and narrative monitoring platform — to analyze how harmful campaigns exploit natural disasters in the U.S.

These narratives do not emerge randomly. They follow repeatable playbooks, leveraging conspiracy-driven claims (e.g., FEMA mismanagement, geoengineering, or land seizures), and are often amplified by super-spreader influencers, bots, and foreign media outlets.

Using LI’s timely detection, narrative clustering, and cross-platform mapping capabilities, we tracked how these campaigns take shape — including the use of AI-generated content, coordinated amplification tactics, and recurring conspiracy themes — and assessed their impact on public trust and institutional response.

Key Findings:

• Recurring Playbooks: Repeated claims of FEMA incompetence, weather manipulation (e.g., chemtrails, Directed Energy Weapons), land seizure plots (e.g., lithium grabs), and hidden casualties.
• Influencer Amplification: Just 5% of accounts generated 40% of total misleading posts — highlighting the role of “super-spreaders.”
• Cross-Platform Reach: X accounted for 52% of detected posts, followed by TikTok (27%), Facebook (14%), and fringe platforms like Telegram and imageboards (7%) — underscoring the scale of cross-platform diffusion (see Figure 1).
• Foreign Influence: State-linked outlets (RT, Sputnik, PressTV, CGTN) seeded narratives hours before domestic influencers amplified them — exploiting disasters to fuel distrust.
• AI-Generated Content: Deepfakes, synthetic imagery, and auto-generated posts accelerated the spread, complicating response efforts.

Case Example – Texas Floods: Within 72 hours of the 4 July flood, LI flagged conspiracy posts (reaching 19.9M views) repeating the geoengineering/FEMA neglect narrative.

Figure 1. Distribution of misleading narratives by platform, based on LI analysis of over 76,000 social media posts related to U.S. disasters.

How Disasters Become Narrative Magnets

Natural disasters devastate more than just physical landscapes — they fracture information ecosystems. In the critical hours after a major event, uncertainty is high and official information is often delayed. This creates an information vacuum quickly filled by conspiracy influencers and adversarial actors.

Misinformation thrives in this vacuum — not just because of its speed, but because it offers emotionally resonant, simplified explanations precisely when confusion and fear peak. From government cover-up theories to fabricated “evidence” of geoengineered storms, these narratives often gain traction before official sources release a single statement.

LI data shows that spikes in online misleading narrative amplification activity closely follow the onset of disasters, with narrative volume typically peaking days after the initial event (see Figure 2). This pattern is consistent across multiple incidents tracked by LI, demonstrating how quickly the information environment becomes distorted.

For example, within hours of the February 2024 train derailment in East Palestine, Ohio, false claims about toxic chemical clouds and bioweapon leaks proliferated across platforms — well before official toxicology assessments were released. This “speed over accuracy” dynamic is now a hallmark of misleading narrative amplification activity during disasters.

Figure 2. Misleading narrative volume before and after selected U.S. disaster events, February 2024 – July 2025. (Source: LI query volume by keyword and date.)

Core Narrative Patterns

Using LI, we analyzed eight major U.S. natural disasters between February 2024 and July 2025 and identified clear, repeatable clusters of misleading narratives. These events included Hurricane Helene, Hurricane Milton, the Los Angeles wildfires, the 2025 Plains Tornado and Wildfire Outbreak, the Texas Panhandle Fires, and the South Texas Floods.

The most dominant themes surfaced consistently across events, often gaining traction within hours of initial impact:

1. FEMA and Government Failure Narratives: False claims circulated that FEMA prioritized aid for undocumented migrants over residents during Hurricane Helene — part of a broader pattern of influence campaigns aimed at portraying federal response agencies as corrupt or negligent.

2. Geoengineering and Weather Weapon Conspiracies: Fake satellite imagery and radar screenshots were circulated online, alleging that Hurricane Helene was artificially created using HAARP or other geoengineering technologies. Circular green patterns attributed to NEXRAD/HAARP interference were falsely identified over cities like Detroit, Chicago, and St. Louis (see Figure 3).

Figure 3. Circulated radar imagery falsely attributing circular storm patterns to HAARP/NEXRAD activity during Hurricane Helene. (Source: X posts, surfaced by LI.)

3. Resource Seizure Conspiracies: In the aftermath of the LA wildfires and Hurricane Helene, viral narratives claimed that evacuations were staged to facilitate government seizure of lithium-rich land or to clear space for "smart city" development projects.

4. Foreign and Domestic Amplification: State-backed media outlets — including Iran’s PressTV and Russia’s RT — amplified FEMA failure narratives during Hurricane Helene, capitalizing on slow federal responses to undermine trust in U.S. institutions.

5. Texas Floods (4 to 10 Jul 2025):  Within hours of flooding in South Texas, HAARP and cloud-seeding theories re-emerged, alongside false claims of diverted federal funds — illustrating how quickly familiar misleading narratives reappear.

Across all eight events, FEMA incompetence and geoengineering conspiracies were the most prevalent and persistent themes (see Figure 4).

Figure 4. Percent of posts coded to four narrative groups: Geoengineering, FEMA/Govt incompetence, Resource sovereignty/“land grab,” and Other (includes hidden bodies, militia calls, and minor themes). (Source: LI)

Texas Floods (4 to 10 July 2025): First-72-Hour Snapshot

Mentions 911 | Daily Average 130 | Potential Reach 19.9 million

In the three days following record flash floods across South Texas and the Rio Grande Valley, LI detected a surge in narrative activity. By 7 July, a TikTok video claiming the storm was an “engineered rain bomb” went viral, increasing daily narrative volume more than fivefold. That same day, RT published an article promoting the same narrative. U.S.-based conspiracy influencers reposted it within hours — a clear instance of the foreign-seeding, domestic-amplification pattern observed across past disaster events.

Narrative distribution for Texas floods:

• 38%: Geoengineering and cloud-seeding claims
• 29%: FEMA mismanagement or federal funding diversion
• 11%: Corporate greed narratives (e.g., BlackRock acquiring flooded land)
• 22%:  Other low-volume themes

As part of its early warning capabilities, LI surfaced the emerging hashtag #TheClearSkiesMovement, appearing in ~600 posts (~3% of the dataset). The hashtag was linked to high-reach accounts within known networks and demonstrated coordinated, cross-platform spread within hours of the disaster’s onset.

Thanks to its timely analytics, cross-platform mapping, and automated alerting, LI enabled stakeholders to act before harmful content reached virality — reducing the impact of misleading narratives in a vulnerable information environment (see Figure 5).

Figure 5. Narrative distribution from 911 misleading information posts during the Texas floods (4–10 July 2025), as detected and categorized by LI in this hashtag cloud.

Amplification and Coordination Tactics

The rapid amplification of inaccurate information relies on increasingly sophisticated techniques — from coordinated bot activity to state-linked narrative seeding — that exploit the chaos following natural disasters.

• Automated Bots and Coordination Tools: Bot networks frequently flood platforms with near-identical information in the immediate aftermath of disasters. Following Hurricane Helene, LI surfaced a burst of anti-FEMA posts within minutes of landfall, traced back to a tightly clustered network of high-frequency, likely automated accounts. By identifying the behavioral patterns and network structures behind these accounts, LI enables teams to quickly assess and respond to inauthentic activity — before it shapes public discourse.
• State-Linked Influence Operations: Foreign state-backed outlets — including RT, Sputnik, PressTV, and CGTN — frequently act as early vectors of divisive or misleading content in the hours and days after a disaster. These narratives are then picked up and amplified by domestic conspiracy influencers, creating a powerful feedback loop that erodes trust in institutions and complicates crisis response.
  
  LI enables teams to detect this pattern early by combining timeline analysis with cross-platform network mapping. During Hurricane Helene, LI surfaced the first coordinated FEMA mismanagement narratives originating from RT (Spanish), PressTV, and RT (English) on 28–29 September — four days after landfall. U.S.-based conspiracy influencers did not begin large-scale amplification until 3–5 October, revealing a clear 4–7 day lag between foreign seeding and domestic uptake.
  
  Figure 6 visualizes this coordinated amplification pattern. On the right, LI identified a tightly clustered group of Russian state media accounts (e.g., @RT_com, @SputnikInt) and affiliates that seeded the narrative early. On the left, LI detected a separate but synchronized cluster of U.S.-based conspiracy influencers and likely inauthentic amplifiers (e.g., @MJTruthUltra), which rapidly elevated the same messaging in domestic spaces. While direct engagement between the two clusters was limited, the thematic and temporal alignment pointed to coordinated — if indirect — amplification dynamics.
  
  By surfacing these clusters and the timing between them, LI enables public sector and enterprise teams to identify high-risk narratives and actors early, prioritize fact-checking or counter-messaging, and preempt escalation before narratives reach virality.

Figure 6. Coordinated amplification network during Hurricane Helene. Clusters reflect Russia state-media seeding on the right and domestic amplification on the left.(Source: LI)

• Recycled and Misrepresented Content: Old or out-of-context visuals — such as disaster images from unrelated events — are frequently used to support false claims, increasing emotional engagement and creating an illusion of authenticity. LI surfaces anomalous content circulation patterns and flags high-traction posts using reused or unverified media — helping content moderation or crisis response teams respond before visual misleading information takes hold.
• Cross-Platform Laundering: Misleading narratives often originate in fringe spaces (e.g., Telegram, TikTok) and migrate to mainstream platforms like X and Facebook, losing source attribution and gaining perceived credibility along the way. LI tracks narrative movement across platforms, correlating hashtags, content clusters, and actor behavior — enabling teams to intercept harmful narratives before they reach broader audiences.

How Logically Intelligence Adds Value

How Logically Intelligence adds value in real time: detects surge in a little more than 45 min; maps cross-platform hops; flags coordinated networks; sends auto-alerts to crisis comms teams.

Identifies early narrative surges, maps cross-platform migration, flags coordinated bot and influencer networks, and issues timely alerts — enabling crisis communications teams to respond before harmful narratives gain widespread traction.

LI equips teams with early insight into emerging disaster-related misleading information — surfacing harmful narratives, detecting coordinated activity, and mapping cross-platform spread. These capabilities support timely interventions that prevent false claims from undermining public trust or crisis response.

The Role of Artificial Intelligence

Artificial intelligence has emerged as a powerful accelerant in the spread of disaster-related misleading information. AI-generated visuals and narratives enable malicious actors to produce high volumes of emotionally manipulative content within minutes of a crisis — long before official channels can respond or verify facts.

Following Hurricane Helene, deepfake videos and synthetic images began circulating within 24 hours of landfall. Among the most widely shared examples was an AI-generated image of a young child in a life jacket holding a wet puppy, allegedly stranded in floodwaters. As shown in Figure 7, this fabricated image went viral with captions accusing FEMA of abandonment and institutional neglect.

These emotionally charged visuals triggered widespread engagement and quickly became embedded in broader conspiracy narratives — including claims of diverted aid, hidden casualties, and government cover-ups. Such content blurs the line between authentic and fabricated imagery, complicating response efforts, eroding public trust, and overwhelming emergency communications teams.

Figure 7. An AI-generated image of a child holding a wet puppy, falsely linked to Hurricane Helene floodwaters, was shared to support claims of FEMA neglect. (Source: Facebook post surfaced by LI)

Why Misleading Information Sticks

Misinformation thrives not only because of its speed but also because it taps into emotional vulnerability, distrust, and simplified explanations during moments of chaos. Sensationalist content tends to spread faster than verified information and is more likely to be amplified by platform algorithms designed to prioritize engagement.

Deep-seated skepticism toward institutions, coupled with the volume and visibility of false claims, allows misleading information to outlast the event that triggered it. These narratives don’t just spread — they stick, re-emerging during future crises in recycled or recontextualized forms.

Using LI, analysts and crisis communicators can track these persistent themes across events and time — identifying narrative patterns like the recurring “missing children” storyline, which has surfaced after the Maui fires, Hurricane Katrina, and Haiti’s 2010 earthquake. Though often baseless, such narratives exploit grief, fear, and a desire for accountability, making them emotionally resonant and persistently reusable. LI’s narrative tracking and clustering capabilities allow teams to detect these resurgences early — enabling proactive response before they shape public sentiment.

Audience Vulnerability and Impact

In the immediate aftermath of a disaster, misleading information frequently spreads faster than verified updates — creating a window of influence that influence actors are quick to exploit. These actors often provide simplistic but emotionally compelling explanations, assigning blame to governments, corporations, or emerging technologies like geoengineering.

In chaotic moments, these narratives can offer a false sense of clarity or control — especially for individuals already distrustful of official institutions. Communities that have historically experienced marginalization or misleading information are particularly susceptible to conspiracy narratives that validate pre-existing fears.

One such example, shown in Figure 8, emerged after the 2025 Los Angeles fires. A viral post on X falsely claimed the wildfires were caused by a DARPA-directed energy weapon to clear land for a “smart city” development. LI’s detection tools traced this post’s rapid spread through a known distrustful online community, illustrating how established audience vulnerabilities can amplify fringe content and fuel broader narratives of government overreach and technological oppression.

Figure 8. A viral X post alleging the 2025 LA fires were a DARPA-directed energy weapon plot to clear land for a future “smart city.” (Source: X, surfaced by LI)

Actionable Recommendations

To mitigate the growing threat of misleading information during disaster response, we recommend the following interventions — each grounded in insights surfaced through LI:

• Deploy Real-Time Narrative Dashboards: Equip emergency responders and public information officers with live dashboards like LI to surface misleading narratives early, track their trajectory across platforms, and trigger timely counter-messaging — before harmful content reaches viral scale.
• Implement Proactive Pre-Bunking Campaigns: Where disasters are seasonal or anticipated, launch pre-emptive, multilingual messaging that addresses recurring falsehoods (e.g., FEMA neglect, HAARP conspiracies). LI’s analysis and trend detection can guide which narratives are likely to reappear and where.
• Engage Early with Known Actors and Networks: At the onset of a crisis, use LI to identify and monitor misleading narrative clusters and high-risk accounts. Early engagement with these networks allows teams to flag content patterns and intervene before narratives spread widely.
• Foster Platform and Government Collaboration: Enhance coordination between social media platforms, emergency agencies, and fact-checking bodies. LI’s alerting and tagging capabilities can support these efforts by identifying high-risk content — particularly synthetic media — for removal or moderation.

Conclusion

Natural disasters create ideal conditions for the spread of misleading narratives. In the absence of timely and verified updates, false narratives — particularly those that are emotionally charged or visually compelling — can quickly fill the information vacuum. As seen in the eight events analyzed, these narratives often coalesce around recurring themes such as FEMA mismanagement, weather manipulation, and institutional failure, frequently gaining traction before official communications begin.

The analysis reveals that such narratives are not only persistent but often driven by coordinated amplification, including state-linked media seeding, automated bot networks, and cross-platform laundering. The emergence of generative AI has further lowered the barrier to producing misleading content, increasing volume, reach, and emotional impact.

Throughout these crises, LI surfaced key signals of narrative escalation, mapped how themes migrated across platforms, and identified clusters of coordinated activity. These capabilities enabled a clearer understanding of the dynamics at play — showing how misleading information spreads, who amplifies it, and how quickly it can influence public perception.

In high-pressure environments where minutes matter, LI’s ability to detect early warning signs, provide cross-platform visibility, and trace coordination tactics offers a meaningful advantage. By equipping crisis communications teams, public agencies, and decision-makers with timely, actionable intelligence, LI can help close the gap between emerging threats and informed response — ultimately reducing harm and preserving public trust.

Sources:

1. Stephen Fowler, “Fact-Checking Falsehoods about FEMA Funding and Hurricane Helene,” NPR, October 7, 2024, https://www.npr.org/2024/10/07/nx-s1-5144159/fema-funding-migrants-disaster-relief-fund.
2. RT, “US Lawmaker Calls for Ban on ‘Deadly Weather Modification,’” RT.com, July 6, 2025, https://www.rt.com/news/621058-deadly-weather-modification-bill/.
3. Yaron Steinbuch, “AI Deepfakes of Hurricane Helene Victims Circulate on Social Media,” New York Post, October 5, 2024. https://nypost.com/2024/10/05/us-news/ai-deepfakes-of-hurricane-helene-victims-circulate-on-social-media/.