Deep fakes are a serious threat and something has to be done to protect people from them, Facebook Inc. Chief Executive Mark Zuckerberg said Wednesday.

Zuckerberg (pictured) said with artificial intelligence getting better, “we need to be very careful” in regards to these fake videos. He was speaking to a crowd on Wednesday at the Aspen Ideas Festival in Colorado.

Indeed, if you heard the recently created Bill Gates’ voice clone or witnessed Zuckerberg himself appearing in a deep fake video, you’d agree that they no doubt will become a big problem.

After that video appeared, Facebook did not take it down from Instagram. The question in terms of policy is when do you take such creations down? When are they just a bit of fun and when are they misinformation designed to hurt or influence? When are they editing, rather than all-out fakes?

“There is a question of whether deep fakes are actually just a completely different category of thing from normal false statements overall,” Zuckerberg said, adding, “and I think there is a very good case that they are.”

Zuckerberg said Facebook is currently in the process of reviewing its policy regarding these videos, although he added that it’s not always easy to define fake. For instance, videos are often edited because the person or people in them want something taken out. A video might be edited by someone else, but should that give the people in them the ability to have them taken down?

Then again, a faked video of a world leader declaring war on another country could obviously cause havoc. “The policies continue to evolve as technology develops,” Zuckerberg said. “But I do think you want to approach this with caution and by consulting experts and not hastily or unilaterally.”

He also said Facebook was slow to take down a deep fake video of House Speaker Nancy Pelosi sounding like she was drunk. He called this an “execution mistake.” This video, he said, “got more distribution than our policy should have allowed.”

Photo: Alessio Jacona/Flickr