The group aims to establish regulations for companies that use copyrighted material to generate images using such text-to-image AI generators as Midjourney, Stable Diffusion, and DALL-E.
The cartoonist Lorenzo Ceccotti and the Italian association of comic book artists MeFu have formed the European Guild for Artificial Intelligence Regulation (EGAIR), an expansion and evolution of their initiative aimed to regulate the use of artificial intelligence.
The group brings together artists, creatives, publishers, and associations from all over Europe who wish to bring to the public attention "how [artists'] data and intellectual properties are being exploited without [their] consent, on a scale never seen before."
EGAIR's goal is to engage with European institutions and establish regulations at a community level for companies that use copyrighted material to generate images using artificial intelligence, including such text-to-image generators as Midjourney, Stable Diffusion, and DALL-E.
"In recent months, while we have worked on the political front of AI regulation, beginning to meet with European Union officials, we have also carried out hard networking work, coming into contact with big names in comics and illustration, associations and trade unions at an international level," MiFu wrote on its Facebook page.
"In fact, we knew that in order to present our requests effectively to the European institutions, it would not have been enough to present ourselves only as a group of Italian authors. EGAIR is our answer to this problem, a point of reference for coordinating European activities and a platform for all those who want to join the cause, support it or find resources and information."
In order to regulate the use of AI, EGAIR has proposed a manifesto whose key points call for:
- explicit and informed consent from data owners before their data is used to train AI models;
- prohibiting the use of names of people or works which are not covered by a license for AI training;
- prohibiting the use of media content not covered by a license for AI training;
- establishing a "human and machine readable" indexing and certification system to report all AI activities and the full content of their datasets;
- the need for a more comprehensive approach to identifying what can and cannot be used in AI learning datasets.
You can learn more about EGAIR and its efforts here. Also, don't forget to join our 80 Level Talent platform, our Reddit page, and our Telegram channel, follow us on Instagram and Twitter, where we share breakdowns, the latest news, awesome artworks, and more.