DENVER — New technology like artificial intelligence is making it easier than ever for scammers to dub people’s voices and likenesses. Scammers can use these deep fakes to call a person’s family and convince them they’re in trouble.
That’s where companies are starting to fight back. Several are beginning to test and roll out their own technology to spot and, in some cases, stop this questionable content.
The Norton Genie is one of those developments. It's a free app that helps people determine in real time if a text, email or social media post is legitimate.
“Basically this tool will be able to help people who lost their trust towards online services to basically check and verify whether a content or a message that they received actually is a scam or not,” said Leyla Bilge, the director of scam research labs at Gen. “It's very easy to use. So, you can only take like a little snapshot of the screenshot of the page that you want to check.”
If the app determines that the content is suspicious, it explains why and offers tips on what to do next.
Norton says within a one-month period, Genie scanned thousands of uploads and determined nearly 45% were legitimate scam attempts.
Each time someone submits a potential scam-- it helps genie's technology to learn so that it can spot future scams with even more accuracy.
“All of us have to kind of contribute to try to solve this problem. maybe this didn't happen to you, but it's going to happen to your kid. it's going to happen to your brother, your sister, your grandmother,” Bilge said.
While scams will likely always be a fact of life in the digital realm, Bilge hopes this technology will help build back some of the public’s trust by empowering them to spot scams.
The company also announced the launch of the Scam Scan challenge, which is an interactive Instagram game where people can test their ability to recognize a scam.
Meanwhile MacAfee recently unveiled its own technology for deep fake audio detection. It's known as Project Mockingbird.
“It's one of our first technologies that we're debuting that analyzes, in this case, audio to assess whether it looks like it was likely generated by AI,” said Steve Grobman, the chief technology officer at McAfee.
While there are a number of legitimate uses for AI having the ability to clone voices, Grobman says there are also a lot of risks. The deep fakes can be used to steal money or even to disrupt the political process. So, McAfee is using AI to fight AI.
“While people have a really difficult time disambiguating a AI-generated audio clone from the original person, artificial intelligence is actually pretty good at detecting some of the fingerprints that the tools used these are a couple examples of how the technology can uncover a deep fake,” Grobman said.
Right now, the technology has a 90% accuracy rate. McAfee is now working on ways to add the tech to its products for customers to use.
In the meantime, these scam stoppers say don't believe everything you see, read or even hear online.
If something seems questionable, look for multiple sources on information and know that if it looks too good to be true it likely is.
