So-called ‘magic avatars’ are the latest internet craze, but one expert says those taking part in the trend may not be aware of the ethics and privacy concerns involved.
Kerry Bowman, a bioethicist at the University of Toronto, says the Lensa app, which creates portraits of individuals using artificial intelligence (AI), has the potential to negatively affect images. one’s self, among other ethical issues.
With Lensa, users can upload 10 to 20 images of themselves to create a “magic avatar,” depicting individuals in a fantasy-like setting where they are said to look “enhanced.” than they really look like.
“It can be exciting in a way, but these idealized images are being driven by societal expectations that can be very cruel and very narrow-minded,” Bowman told CTV’s Your Morning in Monday.
Bowman explains that these descriptions are often unrealistic and do not always accurately depict typical human characteristics. In some examples, people with darker skin tones are shown to be lighter, or people with larger bodies are made to look smaller.
Bowman says these AI programs collect data from the internet in search of different art styles to create these portraits. However, original art styles and artwork created by real artists are rarely compensated for using their work.
“What’s happened with emerging AI is that the law hasn’t been able to really keep up with this in terms of copyright law. It’s been very difficult and very murky and the morals are even behind the law because I think this is what’s going on. It’s basically unfair,” Boman said.
This controversial method Lensa uses is known as the Steady Diffusion model, where many artists don’t have the option to opt in or out of their artwork.
Finally, Bowman said individuals should consider how apps store their personal data and their ability to use their face.
“Do you really want your face to appear in a big database? People need to decide for themselves but it’s not benign, there’s nothing in this, it’s not just a belief. fun,” he said.
According to the app’s privacy policy, photos are only stored in the app for 24 hours; however, it turns out that for each image processed, the algorithm will collect data from the image for better results in the future.
This has led to safety concerns when using facial recognition, Bowman said, especially as this type of data has been contested with police use of it.
Finally, Bowman said individuals should be more aware of their online activity and consider all the potential dangers that come with it.
