Fake images spread fast on Facebook—and its AI filters are missing them

February 23, 2018

There’s a weapon of mass manipulation that Facebook is reportedly struggling to overcome: Photoshop.

Fake pictures: The Wall Street Journal explains that doctored images were a “crucial and deceptively simple technique used by Russian propagandists to spread fabricated information during the 2016 election.”

Pictures are perfect: People love images and share them more frequently on social media than text-only posts. Editing them to inject fake content or misleading messages is a useful way to spread propaganda fast.

This story is only available to subscribers.

Don’t settle for half the story.
Get paywall-free access to technology news for the here and now.

Already a subscriber?

You’ve read all your free stories.

MIT Technology Review provides an intelligent and independent filter for the flood of information about technology.

Already a subscriber?

Hard to block: AI may be able to identify objects in images, but understanding how and why pictures have been altered is more difficult—not least because it requires an understanding of context. And at any rate, Facebook is nervous about censoring content, and doesn’t even automate the process of taking down text posts, which should be easier.

Next up: The Journal says the social network is working on systems that could identify when images are used out of context. But expect fake images to circulate on news feeds for a while yet.

This is your last free story.

Our most popular stories