Author: Charlie Warzel
Publisher: The Atlantic
Publication Year: 2022
Summary: The following article discusses how open source software, which generates images from text, presents two ethical conundrums. First, there is the possibility for abuse by users via the creation of deepfakes. Realistic fake images can serve political purposes, but image generation software can also create fake nude images of celebrities. The second ethical issue is that these tools are usually trained by webscraping millions of images. Many of those images are subject to copyright, and the authors have never been asked for their permission before their work was used to develop the AI.