I recently got a Facebook account to help promote some of my writing and to make a limited amount of connections with friends and family. It’s been a few years since I’ve been on Zuckerberg’s feed of cat videos and embarrassing family photos, unasked-for political commentary and NBA highlights, and so I had the usual suspects in mind after deciding to make the author account. What I didn’t necessarily expect was the blatant amount of AI-generated “slop” on my feed that went rampant after just a week or so of being back on the platform.

I started watching a couple of highlights from the Oklahoma Sooner football team; as a fan, I was interested in seeing some of the highlight reels of the games (needless to say, my heart was broken after Oklahoma’s complete meltdown at the hands of the Texas Longhorns. But that’s neither here nor there.) Soon, however, I started getting questionable headlines and images of football players that were clearly not photographic, but artificial. “Brent Venables sacked as head coach after blowout loss to Texas,” one headline read. Another caption said the quarterback was heading out. One image showed the sullied Big 12 logo on the football jersey, proving the image was both out of date (Oklahoma is now in the Southeastern Conference) and artificially contrived.

Fake Facebook News

It was all fake news, of course. I found myself checking Google news tabs to make sure these claims were false. (Not as if Google is the leader of bias-free news sources either. When you look up “Beethoven” on Google, the main image you get is AI-generated.)

That’s a trivial example of fake news and AI joining forces to mislead a disgruntled football fanbase. However, the AI slop machine holds graver consequences when the images reflect natural disasters, such as the recent Hurricane Helene which ravaged much of the southeast part of the United States in late September.

Fake images of the natural disaster are spilling across social media platforms, particularly on Facebook. Per an article from Futurism, the Facebook page “Coastal Views,” which typically features beautiful natural vistas, has started sharing AI-generated scenes of the hurricane aftermath. Apparently, these AI images are “infiltrating” the Facebook algorithm to attract attention and even money. Maggie Harrison Dupré writes,

Coastal Views’ posts don’t get a huge amount of views, so it’s unclear how much revenue the account is making, if any. Instead, getting Facebook users to the account’s associated Etsy page appears to be the top priority. It’s also not flagging the posts as AI-generated and is instead presenting them as if they’re real images of a disaster zone.

AI companies should be held accountable by being required to watermark their images. In addition, Facebook has to penalize AI-generated content so it’s not inundating everyone’s newsfeeds. The ethical problems of using AI to create lifelike images of natural disasters should be obvious to us. What’s less clear is how to regulate AI images so we don’t always have to second guess a photograph’s authenticity.