Artistic, colourful illustrated selfies have been popping up all over social media thanks in part to Lensa AI.
The photo and video editing app from Prisma Labs, which has been around since 2018, uses selfie photos and artificial intelligence to create portraits in a variety of styles such as anime and pop art. Prisma Labs was founded by Alexey Moiseenkov and a team of Russian developers.
It works by asking users to upload 10 to 20 selfies of themselves, selecting their gender and then asking them to pay to receive 50 avatars (or 100 or 200 depending on price) in different filter styles such as Superhero, Adventure, Rock Star and Cyborg.
Causes for concern
The app is proving popular and the app has skyrocketed to the top of Apple's free apps list, but there have been some concerns over it. One such concern is what Lensa does with photos after they are uploaded. The company claims it does not keep photos.
“As soon as the avatars are generated, the user’s photos and the associated model are erased permanently from our servers," the company said on Twitter. “And the process would start over again for the next request.”
Prisma Labs also says it does not require or request metadata, which can include information like the GPS coordinates of where a photo was taken. But some photos may automatically share that information by default.
While the company's terms and conditions say that users “retain all rights in and to your user content," using the app also grants them “perpetual, revocable, nonexclusive, royalty-free, worldwide, fully-paid, transferable, sub-licensable license to use, reproduce, modify, adapt, translate, create derivative works” of photos.
That means the company owns the rights to any digital artwork made from using selfies in the app. However, for those who are concerned about how their data is being used, they can email email@example.com and request personal data to be deleted. Also, if photos are being used in advertising, users can email firstname.lastname@example.org to remove those permissions. Any approved requests could take up to 90 days to process.
Another problem is copyright concerns as artists have claimed their work is being stolen. It has been noticed that artists’ signatures are sometimes still visible, although scrambled, in some images. The app uses the open source Stable Diffusion model that makes use of copyrighted art from artists around the world in order to work.
Prisma Labs also addressed the issue on Twitter: “The AI learns to recognise the connections between the images and their descriptions, not the artworks,” it said. “This way, the model develops operational principles that can be applied to content generation. Hence the outputs can’t be described as exact replicas of any particular artwork.”
Australian artist Kim Leutwyler took to social media to share that she did a "quick search and found that almost every painting I've posted has been used to train the AI that Lensa is using to create portraits".
"Lensa app is making a profit on stolen, uncredited and uncompensated art."