You may have seen your social media flooded with colorful AI self-portraits over the past week. The viral app ‘Lensa AI’ has been the catalyst behind the influx. Developed by Prisma AI, Lensa turns selfies into what it calls “magic avatars” but has sparked renewed attention recently along with some controversy.

How it works?

Users choose their gender by uploading 10 to 20 images and marking them as “male,” “female,” or “other”, making it inclusive for non-binary people. The program generates the images in 15 to 20 minutes after you choose how many photos you want and pay for them. Avatars can be saved and shared after they are finished.

What’s Stable Diffusion?

The tech behind Lensa AI uses a tool called ‘Stable Diffusion’, an open-source artificial intelligence system that can produce images from text cues, to make its renderings appear hand-crafted rather than computer-generated. There have been rumours that in order to generate these images, other artists’ work have been taken and used which has caused recent controversy.

Privacy concerns?

Of course, like with any artificial intelligence tools, there have been privacy concerns surrounding the app too. While Lensa have stated that they don’t use the photos users provide, the app’s privacy statement only requests authorisation to utilise user content instead of taking full ownership. Images are immediately wiped from servers after generating the portraits.

Want to try Lensa for yourself?

Every year, Lensa costs £36.99. (a one-week free trial is available). Depending on how many photographs you desire, access to the “magic avatars” function costs an additional $4 to $16.