In a webinar this week, the United States Copyright Office (USCO) firmly re-stated its position on artificial intelligence (AI) by stating that it will not register works solely created by AI. This stance aligns AI-generated content with other unclaimable materials, such as previously published or registered works, public domain content, and copyrighted material owned by other parties.
Robert Kasunic of the USCO emphasized that “Human authorship is a precondition to copyrightability,” reiterating the Office’s standpoint. This reaffirms the guidance issued in March, which likened text prompts to instructions given to commissioned artists, drawing parallels to a magazine editor hiring a photographer.
Thomas Maddrey, representing the American Media Society of Photographers (ASMP), shared key takeaways from the webinar. He highlighted that appreciable human authorship remains the primary requirement for copyrightability and expects applicants to become more comfortable with disclaiming materials.
The USCO does not plan to exempt photographers from using the Standard Application, at least for now. The Standard Application requires individual registration and payment per work. However, the Group Application allows photographers to register 750 images at once, paying only one fee for the batch.
How much AI and in what context?
The USCO clarifies its position on works incorporating AI elements during the webinar. According to Petapixel, if AI generates an entire component, such as a forest, it is considered appreciable and must be declared unclaimable. However, if AI only alters an existing element, such as changing the forest colour from blue to green, it is deemed “de minimis” and does not require disclosure.
Photographers planning significant modifications to a photo can register the original version without the AI modifications, even if they intend to display the AI-generated version publicly.
Where do photographers stand with composites?
Despite their reluctance to be labelled as AI photographers, this leaves many artists in a challenging position. Many artists employ tools like Midjourney to create visually captivating works. However, without copyright protection, these creations are vulnerable to unrestricted use by anyone or anything.
Add to that the use of Adobe’s Generative Fill, and you’ve got quite a complicated setup. How different is it from creating a composite image from, say, creative commons images? Is the final result copyrightable?
Legalities around AI are currently in a catch-up position due to the rapid development of the technology. Traditionally it can take several years for new laws to be passed. However, we don’t have that luxury of time when it comes to AI.
[Via Petapixel]
FIND THIS INTERESTING? SHARE IT WITH YOUR FRIENDS!