Local

Experts warn of ‘deepfake’ videos that could trick you ahead of 2020 elections

SEATTLE — Deepfake videos are manipulated to make it look like someone said something they didn’t -- and some of them are alarmingly realistic.

The new technology has experts in the field worried fake videos could be used to influence the 2020 election.

One example is this video of President Richard Nixon announcing to the country the Apollo 11 crew never made it home from the moon.

“Good evening, my fellow Americans. Fate has ordained the men who went on the moon to explore in peace, will stay on the moon to rest in peace,” Nixon says in the video.

“For every human being who looks up at the moon in the nights to come, will know there is some corner of another world that is forever mankind. Goodnight.”

The problem? It’s a speech he never delivered -- words, he never spoke. It was actually created at MIT’s Center for Advanced Virtuality.

Experts there and at the University of Washington are studying the new technology.

“A deepfake is a new family of technology that allows individuals to take videos of people, take their words, and transplant them. Essentially putting words in people's mouths that they may have never said. And it's really difficult when you see these videos, to tell if it's been manipulated or not,” said Jevin West, Directordirector for the University of Washington’s Center for an Informed Public.

He researches and even creates deep fakes.

“It is scary to think about because we can't trust our eyes and ears anymore,” West said.

The technology is something West wants you to be aware of ahead of the 2020 elections.

“Ways of manipulating the words a politician would say before an election -- we'll likely see some of those in our election year,” West said.

West and his colleague, Carl Bergstrom, created a video called “Which Face is Real” to bring public attention to deep fake technology.

One of the two people you see pop up does not exist – instead, they were built with lines of code.

“That we created within seconds with algorithms,” West said. As you might guess, in many of the random parings, it’s very difficult to tell which face is the computer-generated one.

Another example is one that first brought deepfake technology into the public eye in 2017.

It’s a video of President Obama that went viral.

In the video, Obama appears to be saying, “We’re entering an era in which our enemies can make it look like anyone is saying anything at any point and time.” But a split screen shows the person speaking is actually voice actor Jordan Peele.

However, it took months for experts to painstakingly create those “deepfake” videos. Could deepfakes actually impact the 2020 election cycle?

“I think we're there,” West said. “2020 is a year in which we have the technology to confuse a lot of the public, including even people like me who look at these kinds of things all the time, who study them,” he said.

To show you how accessible similar technology is, one app called Doublicat scans a photo of face and puts it on the body of a celebrity of your choice in seconds, spitting out a meme or gif.

The end result isn’t perfect but after all – it’s a free app and took moments to do.

“The technology is getting better and better and it won’t be long before even the experts will have a hard time determining whether it’s a genuine, legitimate video, or whether it’s a deep fake,” West said.

And it’s not just deepfakes. Existing technology that's been around for decades can be used to easily manipulate video and information.

One example is using standard tech to spread misinformation is a doctored video of Nancy Pelosi from 2019. The clip is simply slowed down to make Pelosi sound like she’s slurring her speech.

“It didn't provide a good impression of Nancy Pelosi,” West said. “It was using standard technology of video manipulation and that kind of video can be just as damaging,” he said.

From fake articles, to fake photos, and now -- fake videos -- it's a tough time for consumers to filter all that out and find the truth.

“We live in this world of misinformation and now this technology is making it even worse,” West said.

The goal for West and the Center for an Informed Public is right there in the name -- making sure you're aware of what's out there.

“I don't' think we're going to be able to go back to a world where deepfake technology doesn't exist,” West said. “The more the public can know this exists, that's the best thing I think we can do defensively,” he said.

West said he is working with the Washington state Legislature on a law that would prohibit creating deepfake videos of politicians 60 days before an election. California passed a similar law last year.

But experts say most importantly, if you see or read something really makes you mad or evokes an emotional response, you should question its validity and check what source it’s coming from.

“The technology is dangerous, but it’s most dangerous in these times, when it’s transitioning into the public consciousness,” West said.

He said when Photoshop emerged in the 1990’s, there was a transition period when people didn’t realize images they were seeing had been edited and took it to be truth. But then, the technology became so common that consumers now question images that look “too good” or “too bad to be true.”

“We want the public to get to the point where if they see something suspicious in a video, they’d do the same thing,” West said.