Shortly after a powerful 8.7-magnitude earthquake struck Russia’s Kamchatka Peninsula last Wednesday, July 30, 2025, videos of tsunamis, collapsing buildings and crying children spread online.
The only problem? Many were fake, generated or altered by artificial intelligence (AI). The fake content was created by those hungry for engagement, who took advantage of social media algorithms that enabled its creation.
One so-called tsunami video was from a past event, not an aftermath of the Russia earthquake; another was entirely fabricated. Some people believed them, others were unsure what to believe, but there were still others who panicked. According to news reports, even Elon Musk’s AI tool Grok, embedded in X (formerly Twitter), wrongly told users in Hawaii that the tsunami warning had been lifted when the alert was still active.
In moments of crisis, this misinformation kills trust and has the potential to kill people. When the public cannot distinguish fact from fiction, even responders are caught in the crossfire of doubt.
This isn’t just a Russia or Japan problem. It’s also ours.
We’ve seen AI used for election campaigns, scam messages and even celebrity death hoaxes. But this is different. This is the weaponization of false information at a time when people are most vulnerable. And while AI tools can be helpful for quick access to information and real-time translation, they also multiply the reach of disinformation by bad actors and extend their impact on people in an emergency.
With today’s AI tools of image generators, voice clones and fake news articles, you have a massive misinformation machine.
In places like Cebu or Eastern Visayas, where earthquakes and tsunamis are a real risk, the possibility of AI-generated disinformation disrupting emergency alerts should alarm us. Remember Cebu’s “Chona Mae” tsunami hoax? There was no social media at that time, but imagine if evil people resorted then to creating false information to get engagement. Will the public know better?
Some efforts are underway to address this, from watermarking AI-generated images to fact-checking networks countering false claims and trying to do the debunking fast enough. But government agencies, disaster responders and the local media need to strengthen real-time verification systems, train people to detect fake content and ensure that official communication is more credible and spreads faster than the lies.
We also need to educate the public that NOT everything with a “breaking” label is real. Truth, especially during emergencies, must come first. Not viral content, not AI drama and certainly not misinformation.
The next disaster won’t just test our infrastructure or rescue systems. It will test whether we can still tell the difference between what’s real and what’s AI generated.
Fact-checking and false news debunking should now be part of the arsenal of emergency responders, as any misinformation in the aftermath of a super typhoon or earthquake could lead to lost lives.