Lensa AI Portrait Issue: This App Undresses People Without Permission; What To Do To Stop This From Happening?

It seems that the year 2022 has been a breakthrough year for artificial intelligence, with many new and innovative AI-powered products hitting the market. One of the most popular of these products is a photo-editing app called Lensa AI, which allows users to create anime-inspired digital portraits of themselves, known as "Magic Avatars." These avatars have become incredibly popular on the internet, and their popularity has been linked to that of ChatGPT, a next-generation AI chatbot developed by OpenAI.

In addition to these consumer-facing products, there have also been significant advancements in the field of AI in other areas. For example, text-to-image generators such as DALL-E and Stable Diffusion have disrupted the creative industries, and machine learning has been used to generate fake conversations between living and dead celebrities. There was even an AI rapper that was briefly unveiled and then quickly retired.

Overall, it seems that the advancements in AI technology over the past year have led to a rush to market AI-powered products, and it will be interesting to see what the future holds for this rapidly-evolving field.

Publishing Censored Portraits

The app, which launched in 2018, offers a range of photo retouching features, including the ability to make selfies look prettier by applying filters and removing "imperfections." However, the latest version of the app includes a new feature called "Magic Avatars." This allows users to commission up to 200 AI-generated portraits in a variety of styles by sharing 10-20 images of their faces and paying a small fee. The portraits are generated using Stable Diffusion, an open-source machine learning model that is trained on millions of publicly-available images. According to a report from How-to-Geek, it allows people to mimic different artistic styles, including manga, sci-fi, pop art, and traditional portraiture.

While "Magic Avatars" are a fun and impressive feature, they raise concerns about personal privacy and copyright. Following Futurism's report, this machine learning model is trained on images that are sourced without consent, and there is no way for an individual or artist to opt-out of the dataset. Additionally, the use of AI image generators like this one threatens the job security of professional artists. There is also the concern that someone could create portraits of your face without your permission by using your selfies as input for the AI model.

The popularity of the Lensa AI app and its "Magic Avatars" feature has raised concerns among artists who worry that the mainstream use of AI image generators will make their jobs obsolete. In addition, some people have questioned the cost of the app, which requires a minimum payment of $6 or a yearly subscription of $53.99 to use your photos to create the avatars. There are also concerns that users are effectively paying to train facial recognition AI and giving up their private data.

(Photo : Caitlin Cassidy/The Guardian)
Photograph of Guardian Australia reporter Caitlin Cassidy and an AI-generated image based on several selfies sent to the Lensa app’s ‘magic avatar’ generator. Lensa, an AI art generator, flooded social media with celestial and anime-inspired "magic avatars" which sparked discussions on the ethical implications of using the app.

ALSO READ: Tesla Malfunction: Full Self-Driving System Seen Violating Traffic Signs on School Buses

User's Privacy

The app's creator, Prisma Labs, has previously gotten into trouble for accidentally generating nude and sexually explicit images using the app, despite having a "no nudes" and "adults only" policy. The company's CEO has said that this behavior only happens if the AI is intentionally provoked to create this type of content, which is against the app's terms of use. However, some have expressed concerns about the potential for misuse of the app and the impact it could have on users' self-image and body image, as stated in the TechCrunch report.

In response to these concerns, Prisma Labs has stated that it does not use user photos for any purpose other than applying stylized filters or effects, and that face information is automatically removed after 24 hours of being processed. The company also emphasized that users' images are only used to create their avatars.

In addition to the risk of nonconsensual pornography, there are also concerns about the potential for AI to be used to create political misinformation and disrupt education. Overall, the Lensa AI app serves as a reminder that AI technology is still in its experimental phase and can have unintended consequences if not properly regulated.

The use of AI art generators to create non-consensual images, particularly those of a sexual nature, is a growing concern. The ease with which such images can be created and the potential for them to be widely distributed raises serious ethical questions. The UK government has responded to this issue by proposing laws that would criminalize the dissemination of non-consensual nude photos. While this is a step in the right direction, the challenge of governing the internet and addressing the complex moral and ethical issues surrounding AI art generators remains significant.