Emory professor believes AI could be a tool to fight against AI Robocalls

Of particular concern, are people using AI to generate voices that sound like loved ones to scam victims out of money.

In July, we brought you a story of Debbie Shelton Moore. She picked up a call from a local number and heard what she thought was her 22-year-old daughter crying.

“She’s saying ‘Mom, mom…and it’s just so authentic’,” she told Channel 2 Action News. “The scammers said ‘She’s been kidnapped, we have her in the back of the truck, and we need $50,000′.”

Moore’s husband called 911 after they pulled up an app on her phone showing her daughter’s location and then called her.

It’s phone calls like the one the Moores received that have the White House promoting a “hackathon” to see if they can fight fire with fire. Or, in this case, AI with AI.

“Currently, we can do 80 percent of the detection of AI with AI,” says Emory professor Rajiv Garg.

Garg says AI has the potential to fight back. He believes the technology can better identify the pauses and other details that separate artificial from human speech that humans can.

TRENDING STORIES:

• Squatters hopping from home to home across metro Atlanta, armed with fake leases
• Atlanta rapper T.I. celebrates opening of his first affordable housing development
• There is going to be a big change in the weather by tomorrow morning

“The AI has to catch up and understand every nuance in the sound wave,” Garg said. “The issue is it is getting better, it’s getting increasingly more realistic.”

Garg says defending against robocalls and AI is harder than creating attacks.

“Instead of 30 lines that were used to create an attack, now I need 80 lines and a much faster machine to detect it,” he said.

However, he believes the technology can work if experts in the field start coming up with solutions.

“If we start now, we have some hope to protect people,” said Garg.

[SIGN UP: WSB-TV Daily Headlines Newsletter]

IN OTHER NEWS: