The clips on Manjeet Rege’s cell phone might look real — but they’re actually fake.

“It is scary, and it is important that everyone learns about AI’s technology and power,” he said.

In those videos, Rege, a Data Science Professor at the University of St. Thomas, appears to be speaking in Chinese, French and Hindi.

The professor, who isn’t fluent in any of those languages, created the fake videos himself.

“So, consumers need to know that what you see should not be believed right away,” he noted. “It could be something that looks authentic but is not.”

It’s a lesson that Rege will soon be teaching at the university — a master’s degree course in artificial intelligence.  
 
“We want to teach our students not only the technology behind AI,” he said. “But also, now that they have access to so much technological power, to deploy that in an ethical manner.”

Experts say there’s already a huge demand for people who can develop and implement AI applications.

The skill sets could range from inventing new products or services to detecting cyber-attacks on a company’s database.

“How can this be used by a bad guy, try to figure out the bugs before the bad guys find them and abuse them,” explained Brian Halbach.

Halbach, who runs a Minneapolis cyber-security firm called ‘Good Guy Hackers,’ says it’s increasingly important for companies to have an AI expert on board.

“I think it’s another play of good versus evil,” he said. “You’re going to need a good AI person to integrate into what you’re doing, and also protect your intellectual property.”

The notion of educating AI specialists comes with growing worries about bad actors and deep fakes.

Even the White House is voicing concerns about social media platforms after fake and explicit images of Taylor Swift went viral on the internet.  

“It is alarming,” declared White House Press Secretary Karine Jean-Pierre. “We believe they have an important role to play in enforcing their own rules to prevent the spread of misinformation and non-consensual, intimate imagery of real people.”

The Biden Administration says Congress should enact legislation, calling for the creation and sharing of those images a federal crime.

Minnesota already has a deepfake law, passed last year, which prohibits the non-consensual sharing of sexual deepfake images, or using deepfakes to interfere with an election 60 days before polls open.

Violators could face up to five years in prison and $10,000 in fines.

“We need to have some regulation to control the application of how AI is being used, so it does not go into the wrong hands,” Rege said. “In every technological advancement, there are both sides, the good and not-so-good. We hope we’re teaching that side to our students as well.”

Both Rege and Halbach say they hope AI education, including the ethical use of technology, will make a difference.

“Hopefully there’ll be enough good AI researchers who can pump the brakes on some of the unethical work that’s going on out there,” Halbach said. “That would help out.”