Real Spain flood photos are getting flagged as AI fakes

A harrowing photograph from the catastrophic floods in Valencia, Spain, stopped everyone in their tracks this week. It showed around 70 cars tossed like toys down a narrow street, wedged at bizarre angles after an eight-hour deluge dumped over 490 liters of rain per square meter.

Then came the internet's collective verdict: fake news, entirely generated by Midjourney.

It was not fake. Technology journalist Charles Arthur tracked down a bar sign visible in the frame, located the exact street on Apple Maps, and verified the scene via Street View. But the uncanny sheen of the cars, the compression of the telephoto lens, and the sheer absurdity of the disaster looked so polished that millions simply refused to believe it was real.

Feeding the engagement beast

People didn't doubt the photo because it looked poorly made. They doubted it because social media feeds are drowning in what researchers call AI slop, conditioning everyone to suspect reality itself.

If you thought tech platforms were trying to fix this flood of low-grade synthetic content, think again. According to reporting from 404 Media, Meta CEO Mark Zuckerberg told investors on a recent earnings call that AI-generated feeds are actively coming to Facebook, Instagram, and Threads. Zuckerberg framed AI-created and AI-summarized content as a major product opportunity.

The math is straightforward: platforms need attention, and synthetic content is practically free to manufacture.

On the supply side, content farms have already industrialized this. Investigations revealed creators in India generating stomach-turning or emotionally manipulative AI pictures, like frail elders swarmed by insects, solely to harvest platform creator bonuses. In some setups, 1,000 likes can yield up to $100.

Creators profit from the bonuses, platforms profit from the raw watch time, and the feeds grow progressively more unhinged.

Why it matters

We are watching the flip side of the deepfake panic play out in real time.

For years, experts warned that generative AI would fool people into believing fabricated events. The Valencia photo proves the inverse is already happening: synthetic media has become so ubiquitous that people reflexively dismiss authentic human tragedy as a digital hallucination. Real disasters lose their urgency because viewers assume an algorithm cooked them up for cheap engagement.

For everyday internet users, reality verification now requires private investigator skills. Nobody has time to pull up Apple Street View just to confirm an evening news photograph, which means widespread apathy is inevitable.

Meta's push to officially bless AI-generated feeds looks especially reckless. By treating synthetic junk as a legitimate engagement driver, social networks are accelerating their own decay. When an ecosystem rewards cheap machine slop and disincentivizes actual documentation, it erodes the final shred of trust holding digital public squares together.