AI-generated Flood Content Poses New National Security Threat, Risks Panic, Say Experts

KOTA BHARU, Nov 30 (Bernama) -- Increasingly realistic Artificial Intelligence (AI) generated videos are a key factor behind the rise in fake content being circulated during flood disasters, creating public confusion, risking widespread panic and posing a threat to national information security.

Dr Noor Nirwandy Mat Noordin, Warfare and Security & Political Analyst at Universiti Teknologi MARA (UiTM), the current Northeast Monsoon season shows the public remains highly susceptible to deceptive AI-generated videos and images.

“We are now facing a new era where anyone can produce highly convincing videos or visuals related to floods, storms, or any disaster. If left unchecked, this can lead to information chaos,” he told Bernama today.

He explained that while AI innovation brings many benefits, such as creating infographics and disaster awareness materials, the public must exercise 'self-stoppage', a form of self-regulation to verify information before sharing it. 

He added that the surge in fake content is also driven by social media users who unintentionally trust AI-generated visuals, thereby contributing to the creation of false narratives and perceptions, especially during a crisis.

“Therefore, the practice of ‘tabayun’—pausing and verifying information—must become a cultural norm. The public should utilise fact-checking platforms such as sebenarnya.my, MyCheck and CheckFacts to ensure the information they receive is genuine,” he said.

He also cautioned that the threat of AI-based information manipulation can affect public sentiment, influence their responses, and undermine public order as well as national security.

“When panic occurs due to unverified information, rescue coordination can also be disrupted. This is why we need a psychological mechanism for information warfare: the public's ability to assess information rationally and not be swept away by manipulated content,” he explained.

He added that Malaysia should consider establishing a national disaster information centre and a dedicated fact-checking segment for disasters to ensure a more organised, rapid and accurate flow of information.

“A national disaster calendar could also be developed as a periodic reference to help the public prepare, while security agencies focus on strengthening rescue coordination,” he said.

He said every level of society must now become a "bastion of information defence" to avoid being easily manipulated by extreme or fake AI content, especially during a disaster.

Echoing this sentiment, Crime Analyst Shahul Hamid Abdul Rahim confirmed that this unhealthy practice has now been identified as a new threat to emergency management, as generative technology can now produce realistic flood visuals in a matter of minutes.

He said this phenomenon has given rise to what is considered an 'information disaster', where the impact of a crisis is no longer confined to physical damage but also includes information disruption, which can potentially trigger panic, confusion and interfere with official communication.

He added that this problem is expected to worsen so long as the level of digital media literacy among the public remains low.

"The dissemination of fake visuals can also disrupt rescue operations by the fire department, the Civil Defence Force and the police. Resources may be diverted to locations that appear affected in a video, while genuinely critical areas are overlooked. This increases the risk of wasted resources, delayed responses and coordination errors," he said.

He added that emergency hotlines also risk being overloaded when the public makes repeated calls to verify the authenticity of a video, meaning critical calls may go unanswered.

"The spread of information that contradicts official statements also weakens public trust in such agencies and volunteer teams," he added. 

According to him, fake AI visuals can also be exploited for fraud, including fake donation appeals, the exploitation of public sympathy and the manipulation of emotions, adding a new dimension to digital security threats.

"From a technical standpoint, the public finds it increasingly difficult to distinguish real content from AI-generated videos, as the quality of visuals becomes more realistic, complete with weather effects, water, mud and sound.

"A lack of digital literacy training, the public psychology during disasters that makes people believe dramatic content, and social media algorithms that prioritise sensational material all contribute to fake videos spreading faster than official information," he elaborated. 

Shahul Hamid added that the absence of a rapid-reference centre to verify disaster-related content also leaves the public vulnerable to deception.

"Creating and sharing fake AI videos during a disaster is a form of technological misuse and a violation of public communication ethics. Such content can also be exploited by certain parties to create narratives of distrust towards the government and rescue agencies," he explained.

Meanwhile, a survey by Bernama on TikTok and Instagram found a growing number of AI-based videos appearing in the wake of the Northeast Monsoon season, affecting several states.

The detected fake content includes visuals of crocodiles emerging, tigers entering villages, venomous animals, and various extraordinary situations portrayed as happening during floods. 

Also going viral are videos of major floods that look highly realistic, leading some users to believe they are genuine field recordings, when in fact they are entirely generated by AI technology.

The ability of AI to produce visuals that closely resemble real events is causing significant confusion among users, particularly those less familiar with the characteristics or tell-tale signs of manipulated content.

When shared unchecked, these videos cause public anxiety and misrepresent the flood situation, creating misunderstandings and distorting reality.

-- BERNAMA