“Deeper Fakes”

Artificial Intelligence has sparked mounting concerns as it is being used more and more often in political spheres. 

About the democratization of AI, Associate Professor of Communication at Pitt-Greensburg Jessica L Ghilani, says, “A lot more people have the ability to do this kind of work in ways they couldn’t before, and that’s where I think it gets ethically murky,” 

“Deepfakes”, a form of AI using video and imaging, have done such a convincing job at creating life-like images, that people are finding themselves unable to distinguish between computer-generated images and real ones. When utilized in the political arena, they have the potential to be highly dangerous.

“Deepfakes” have been around for a long time in movies and television, but more recently they have been used during elections to digitally force a candidate to look a certain way or to have them say something they wouldn’t speak otherwise.

“Deepfakes make our trust in visual information compromised,” says Ghilani.

With the recent indictment of former President Donald Trump, there has been an influx of AI creations in his likeness. Some artificially-generated images portray a mug shot with his face on it, and others show an image of the former President being led by police on a perp walk that was never actually caught on camera.

“I think most people encounter it in ways that seem innocuous, but the risk comes in terms of foreign policy, in terms of our political realm, that’s where these videos can potentially do harm,” says Ghilani.

The ability to spot deepfakes is getting increasingly difficult as technology progresses. Twenty years ago, rudimentary deepfakes were easy to point out. Nowadays, the general public has a harder time differentiating.

“I think the best way to identify it is to look at the encoded language of the video,” says Ghilani.