In Facebook, videos that are viewed by thousands of people show CNN's Wolf Blitzer seemingly promoting a diabetes drug. In another video, "CBS Mornings" host Gayle King appears to be endorsing a weight loss product.

However, these videos have been manipulated - this is the latest case of a large number of deepfake images, which use the images of trusted news personalities and place them in false advertisements, undermining people's trust in the news media.

In recent months, similar social media posts have used the images of Fox News personality Jesse Watters, CBC host Ian Hanomansing, and BBC stars Matthew Amroliwala and Sally Bundock, among many other well-known figures.

Many of the exploited news personalities have used their own accounts to refute these claims.

King said on Instagram in October, "I have never heard of this product nor have I used it!"

After seeing a video clip that appeared to show him promoting marijuana products, CNN's medical correspondent Sanjay Gupta also issued a warning, saying, "These scams have nothing to do with me... I am primarily concerned about your health and I am genuinely worried that you may be harmed by using these products."

These manipulated videos promote unverified treatments and investment schemes - promising "guaranteed returns" or the opportunity to obtain coveted shares. Some videos also use altered footage of Tesla and SpaceX billionaire Elon Musk.

Some of the videos include links to investment schemes, unapproved products, or unrelated e-commerce websites, which disappear after a few days.

Meta, the parent company of Facebook and Instagram, has banned deepfake content since early 2020, with exceptions for parody and satire. Other platforms have similar policies.

However, many similar clips continue to spread online.

Voice Cloning

"I have seen an increase in this type of video, where a person's voice is cloned and with just two minutes of their voice, any other video of theirs can be modified to match the mouth movements with the new audio," said Hany Farid, a professor specializing in digital forensics at the University of California, Berkeley.

Some deepfake works are easily detected due to their poor quality. However, experts warn that the technology is improving - television personalities are easy targets because they have a large amount of footage available for training AI programs.

This trend is concerning because "people have become accustomed to trusting news anchors like they trust their friends," according to Andrea Hickerson, dean of the School of Journalism at the University of Mississippi.

She said, "This is very dangerous because people don't expect misinformation and disinformation to come in this way. It looks like traditional news media."

"Trust Crisis"

AI-manipulated content has become an increasingly significant part of investment fraud, resulting in approximately $3.8 billion in losses for American consumers in 2022.

These scams have targeted victims in Canada, Australia, and other countries. In some cases, individuals have lost tens of millions or hundreds of thousands of dollars.

"As criminals combine traditional tactics with online scams involving cryptocurrency and artificial intelligence, the scams are becoming increasingly sophisticated," said lawyer Chase Carlson in a blog post earlier this year.

Americans are growing increasingly concerned about the use of AI online, especially when it comes to politics.

According to a survey conducted by Axios and business intelligence company Morning Consult in September, over 50% of people expect these lies to impact the results of the 2024 elections.

Rebekah Tromble, director of the Institute for Data, Democracy, and Politics at George Washington University, said this type of misinformation "raises broader concerns about trust in information and institutions."

According to an October Gallup poll, only about one-third of Americans have "a great deal" or "quite a lot" of confidence in the news media, matching the lowest level in 2016.

Tromble noted that many manipulated clips circulating online are low-quality "cheap fakes," but they still contribute to a "trust crisis." She urged news consumers to be cautious before sharing these posts on social media.

"However, it also brings some benefits - with a healthy dose of skepticism, we can identify what is misinformation," she said.