Artificial intelligence and ‘deepfakes’ could spread life threatening misinformation in emergencies

RICHMOND, Texas – When a hurricane is eyeing up the Texas Gulf Coast, we all want to know the most up-to-date information every single minute.

Technology has given us the resources to do just that. But at the same time, that same technology could be used to spread misinformation just as fast as real updates.

Social media started a fire of misinformation, allowing anyone to post just about anything. It could be true or it could be false.

Now, with advancements in artificial intelligence, it’s becoming harder to sort through what’s fake and what’s real.

The introduction of Deepfakes just poured jet fuel on that fire.

Defining Deepfakes

Deepfake (n) – an image or recording that has been convincingly altered and manipulated to misrepresent someone as doing or saying something that was not actually done or said

That’s the official definition from the Merriam-Webster dictionary.

To learn a little more, we lean on the experts at the University of Virginia.

“A deepfake is an artificial image or video (a series of images) generated by a special kind of machine learning called “deep” learning (hence the name). There two overviews of how deepfakes work in this article: one for the layperson, and one for the technically-minded,” the university shared online. “Deep learning is similar to any kind of machine learning, where an algorithm is fed examples and learns to produce output that resembles the examples it learned from. Humans learn the same way; a baby might try eating random objects, and it quickly discovers what’s edible and what isn’t.”

You Might Use Deepfake Technology Everyday

That’s right. The technology that fits in the palm of your hand and lives in your pocket is the same type of tech behind deepfakes.

Apps like Face Swap, filters on Instagram and Snapchat and apps that alter your voice or allow you to type text and have your voice say the words are all examples of the machine learning that’s used to create deepfakes.

“They kind of do it now. It’s kind of like a joke,” said Ariana Elias of Stafford.

The difference is the complexity of the machine learning.

A simple app like Face Swap doesn’t use a lot of resources.

Meanwhile, creating a video of someone saying something they never did is a very resource intensive process.

Deepfakes During Dangerous Situations

Deciphering between a deepfake and what’s real can be really difficult. And here’s the real problem: it’s only going to get harder.

During an emergency situation, like a hurricane or other natural disaster, taking the time to analyze a piece of information, for example a statement from a press conference held by the local emergency management office, could mean evacuating before a storm hits or staying put.

“I am actually really, really concerned about that on many levels,” said Roderi Holmes of Stafford.

It’s that exact fear that presents a new challenge for Fort Bend County Emergency Management Coordinator Greg Babst.

He’s no stranger to the danger deepfakes pose to the community. But it wasn’t until a recent training conference that he first hand got to experience a deepfake of himself.

“One of the cyber analysts came in there and they basically took my information,” Babst explains. “During the end of the conference. They were able to put up their presentation and using AI and only an hour of time, that person was able to grab my face off of social media, was able to grab my voice over from press conferences and whatnot that I’ve done in the past on social media from our sites, and then put that capability with AI and putting me in an emergency operation center and telling people to evacuate.”

It’s that very experience that opened a whole new vulnerability to getting life-saving information out fast, but also accurately.

Gage Goulding: “Was that experience eye opening for you?”

Greg Babst: “Yes. I honestly knew it was out there. I didn’t know that it could be that almost that real.”

Don’t Be Afraid, Be Aware

During a time of emergency, a deepfake video of someone like Babst, a mayor, governor or county judge could put potentially life-threatening or deadly misinformation out into the world.

You shouldn’t be scared of the world, but instead don’t take everything at face value until you investigate the source and ensure it’s coming from a trusted, vetted place.

“Know your sources, vet those sources and then continue to follow those exact sources,” Babst said.