Hurricane Helene Is Causing Serious Damage. However, Most of the Photos of its Destruction on the Internet Are AI-Generated

The photos above are shocking. In them, a girl wearing a life jacket cries inconsolably while hugging a small puppy aboard a boat. All around her is what looks like a flood caused by some natural disaster.

They may look like real images, but they’re not. They’ve been generated by artificial intelligence.

As reported by the newsletter Garbage Day, the images went viral on Facebook and then were shared on other social media platforms by X and YouTube users, including supporters of former President Donald Trump.

Users who spread the images are taking advantage of the destruction caused by Hurricane Helene for political purposes. They’re criticizing the efforts of the Federal Emergency Management Agency and spreading images and fake news to undermine the rescue and relief operations for those affected by the tragedy.

It’s almost impossible to find the source of these and other images. In fact, it’s proving to be extremely difficult to stop the spread of false information about Hurricane Helene. Facebook has added tags to these images and others to indicate that they’re an “altered photo.” The social media company is also advising users to verify the photos’ authenticity before sharing them.

To make matters worse, users have shared the fake images on pages using the hashtag #NorthCarolina, one of the states affected by the hurricane. The same pages, which many affected people are likely consulting, were previously providing useful and legitimate information about the disaster. Now, they’re being filled with conspiracy theories.

For example, one theory claims that the flooding wasn’t caused by the hurricane but by human action related to lithium deposits and AI development companies. Another theory says the government is to blame for the North Carolina incidents.

The concerning issue is that, despite numerous warnings, people continue to share AI-generated images even when they know they may not be real. Republican National Committeewoman Amy Kremer was also duped by the fakes, posting an image on X that was later flagged as a deepfake in a community note. However, she declined to delete her post with the image, stating that its authenticity didn’t matter and that “it is emblematic of the trauma and pain people are living through right now.”

This isn’t the first instance of similar behavior on social media, and it likely won’t be the last. The problem lies in the increasing believability of fake content and the lack of effective measures by platforms like Facebook and X to control or prevent it. While Community Notes may help, they often come too late. Deepfakes have become a powerful tool for shaping public opinion, and current platforms are ill-equipped to combat this issue.

Related | New Tool for Creating Deepfakes for Free Rekindles Debate on the Dangers It Poses