Facebook uses AI to detect fake ads

They’re stepping up their enforcement of a technique called “cloaking.”

Getty Images

Facebook has been making efforts to be more transparent about its efforts to moderate posts and combat the spread of fake news and other spam, especially after the social network's effects on the recent election in the US. Today, the organization disclosed how it is using AI to help combat the issue of "cloaking" in Facebook News Feeds.

When you click on a post or an ad in your News Feed, the target site should match what the ad is about -- it's pretty straightforward. But sometimes, it doesn't. And in fact, the target site can even violate Facebook advertising policies and/or community guidelines. This is where cloaking comes in; the ads and posts are able to get around Facebook's review processes because they send Facebook's employees and contractors to different sites than users clicking in their News Feed.

Facebook has been putting more resources into its detection strategies, using a combination of AI and human review, to find and remove cloaked posts and ads. Additionally, they've amended their policies to include strong language that forbids cloaking. Any advertisers or Facebook Pages that are found to use this practice will be banned from the social network. It's a small step in combatting spam, but certainly a welcome one.