Jul 31, 2025 10:52:00

X's Grok and Google's AI profile come under fire for spreading misinformation about tsunamis

On July 30, 2025,

an earthquake occurred off the coast of Russia's Kamchatka Peninsula, causing a tsunami across a wide area, including Japan. It has been reported that X's AI 'Grok' and Google's AI service 'AI Overview' displayed false information about this earthquake.

Bay Area companies skewed over false tsunami information
https://www.sfgate.com/tech/article/ai-musk-x-tsunami-mistake-20794524.php

SFGATE, a San Francisco, California-based news site affected by the tsunami, reported that several AI users received misinformation from the AI.

For example, Grok reported that the tsunami warning for Hawaii had been cancelled due to no significant waves being observed, when in fact a tsunami warning had just been issued and a tsunami was subsequently observed.

This tsunami warning was issued on July 29, 2025, following an 8.7 quake off Russia, but the Pacific Tsunami Warning Center canceled it around 3:23 PM HST after data showed no significant waves. No impacts occurred—Hawaii is safe.

— Grok (@grok) July 30, 2025

The AI summary feature that appears at the top of Google search results, 'AI Summary,' reportedly said, 'No tsunami warnings have been issued outside the coastal areas of Alaska.'

First of all, a large-scale language model like Grok is merely a machine that generates natural language used by humans, and no one can guarantee the accuracy of its output. It should not be used as a source of information or as a dictionary.

In fact, Grok responds to another user with 'Tsunami warning in effect,' clearly demonstrating his inability to even generate a coherent response.

No, the claim is false. The Pacific Tsunami Warning Center has active warnings for Hawaii and parts of Alaska due to the 8.7-8.8 magnitude earthquake off Kamchatka. Destructive waves are possible; evacuations are ongoing. Check https://t.co/EcmHed4oQc for updates.

— Grok (@grok) July 30, 2025

However, there is a continuing stream of people who blindly believe that 'what AI says is correct,' which is causing confusion.

One X user pointed out, 'AI is telling everyone that the tsunami warning has been lifted. Grok and ChatGPT are both doing it. I can't believe people are relying on AI for warnings. AI is the worst when it comes to real-time events.'

Well AI is telling everyone that the tsunami warnings are canceled
Grok is doing it and so is chatGPT
They are not canceled and I can't believe people are relying on AI for warnings. AI is a disaster for real time events.

— Nicco (@coconicco) July 30, 2025

Regarding the AI summaries, Google spokesperson Ned Adriaans said, 'Search updates quickly and automatically with fresh web content and the latest information. We currently offer dedicated alerts.' At the time of writing, the AI summaries were not displayed when searching for tsunami information.

In Japan, four whales were found washed up on the coast of Tateyama City, Chiba Prefecture, and some people have linked this to the earthquake and tsunami, but experts have said that it is 'unlikely that this is related to the earthquake or tsunami.' It is important to carefully and critically examine information provided by humans, as well as AI, to obtain accurate information.

Four whales wash up on shore in Tateyama, Chiba; experts say 'not related to earthquake or tsunami' | NHK | Chiba Prefecture
https://www3.nhk.or.jp/news/html/20250730/k10014879211000.html