Here are some worrying findings from a new study by Lancaster University in the UK and the University of California, Berkeley:
Fake faces created by artificial intelligence (AI) are considered more trustworthy than images of real people.
The results revealed that synthetically generated faces are not only highly photo-realistic but de-facto indistinguishable from real faces and are even judged to be more trustworthy.
The researchers call for safeguards to prevent "deep fakes" as people cannot distinguish between a face generated by Artificial Intelligence—using StyleGAN2—and a real face.
How exactly did they study this?
- The research experiment asked participants to identify a selection of 800 faces as real or fake and to rate their trustworthiness.
- After three separate experiments, the researchers found the AI-created synthetic faces were, on average, rated 7.7% more trustworthy than the average rating for real faces.
If you want to give it a try yourself, take a look at the collection of images below.
Some are real, some are fake. Can you tell the difference?
What makes this generation defining?
Well, many of the deep fake examples around right now are simply fun parodies or experiments designed to test the limits of deep learning technology.
Maybe you have seen the TikTok account dedicated entirely to Tom Cruise deep fakes? If not, check it out right here: https://www.tiktok.com/@deeptomcruise?lang=en
Here is one example:
- Last year, the Korean television channel MBN presented viewers with a deep fake of its own news anchor Kim Joo-Ha.
- Based on technology from DeepBrain.ai, MBN now uses the deep fake quite regularly for breaking news reports.