Mergen Demo Gadget Lab 38/60
Skip to content
AI image generation: The debate over copyright, ethics and labeling
AISoftware

AI image generation: The debate over copyright, ethics and labeling

Images generated from a few words are affecting many professions, from illustrators to photographers. We heard from every side of the debate.

SLSophie LaurentArts & Entertainment Editor
9 min8.1B0
Table of contents
  1. How do image generators work?
  2. Artists' copyright concerns
  3. Should AI content be labeled?
  4. Key takeaways

Text-to-image AI tools can create striking illustrations and photorealistic images in seconds. While these tools speed up designers' work, they also raise copyright concerns among artists whose works were used in training data without permission. Illustrator Natalie Evans, whom we spoke with, describes her disappointment on discovering that her style had been imitated by AI. Legal experts, meanwhile, say existing copyright rules do not fully address this new situation. Clearly labeling AI-generated content stands out as the point of greatest agreement in the debate.

How do image generators work?

AI image generators turn a sentence you type into a picture in seconds. Behind the scenes is a model that has learned from millions of image-and-caption pairs. Rather than copying an image, the model learns the statistical relationships between words and images and creates a new one. Even so, the output can resemble the training data, so it isn't right to consider every output entirely original.

Designers use these tools to sketch out concepts, prepare presentation visuals and generate ideas. The speed gain is huge: ten different visual takes on a product idea can appear in a few minutes. On the other hand, output quality depends heavily on the prompt and on repeated attempts; hands, text and complex scenes still produce errors.

How do image generators work? — AI image generation: The debate over copyright, ethics and labeling

What a tool can do doesn't remove our ethical responsibility for how we use it.

Sophie Laurent

Artists' copyright concerns stem from where the training data comes from. Many models were trained on images collected from the internet, and the owners of those images were neither consulted nor paid. Illustrator Natalie Evans describes her disappointment on realizing that the drawing style she spent years developing could be imitated with a few words. The debate has an ethical dimension as well as a legal one.

According to legal experts, current copyright rules don't give clear answers for this new situation. Whether a style can be protected, whether using training data requires permission and who owns the generated image are all contested. Some companies are charting a new course by offering models that pay participating artists or don't use copyrighted content.

Should AI content be labeled?

Labeling is the point of greatest agreement in the debate. Clearly stating that an image was generated with AI, especially in news, advertising and political content, keeps readers from being misled. Some platforms have started using hidden markers that document the content's origin, along with visible labels. That's why experts stress the habit of verifying sources alongside labeling, and expect readers to stay alert.

Topic AI
Highlights AI, copyright, ethics
Reading time 9 minutes
Written by Sophie Laurent

Key takeaways

  • Image generators learn from millions of examples to create new images.
  • Artists worry about training data used without permission and style imitation.
  • Clearly labeling AI content is the broadest point of agreement.

Labels, however, can easily be cropped out or removed, so they aren't a solution on their own. Media literacy, the habit of questioning sources and regulation must move forward together. What a tool can do doesn't remove our responsibility for how we use it. In our own publications, we clearly mark every AI-generated image. We also encourage our readers to make questioning an image's source a habit.

Share this post

What do you think about this post?

SL
Arts & Entertainment Editor

Sophie Laurent

A former film, music and performing arts reporter, Sophie Laurent covers festivals and exhibitions alongside the colorful world of entertainment news. In her interviews, she likes to put the story the artist is telling front and center, rather than the artist.

Leave a comment

Share your thoughts

Your email address will not be published. Respectful, on-topic comments are approved.

Gadget House

Welcome back

Log in to your account; your saved posts and comments are waiting for you.