AI Art Generation and Copyright Law

In cases where there is NO HUMAN INVOLVEMENT, Urantia Found. v. Kristen Maaherra, 114 F.3d 955, 957–59 (9th Cir. 1997), has indicated that the work is solely created by AI and no copyright will issue.

The newest spin on these AI issues exists when the Artist feeds his underlying work to a computer and uses AI to apply styles it has “learned” by scanning named Artists like DaVinci, Michaelangelo, etc.

Andersen et al v. Stability AI Ltd. et al, Case No. 3:23-cv-00201 in the United State District Court for the Northern District of CA, San Francisco Division, is a class action suit wherein several artists are suing AI art generators Stability AI, Midjourney and DeviantArt for using their work to train AI tools. The suit contends defendants downloaded billions of images from the internet without consent which were used to create AI styles of the artists. Purchasers of the technology can generate new works in the style of the named artist without compensation.

Is the scanning of art and creating a style which is applied by a computer to an original photo to make it a work of Art with “learned” techniques of what was scanned an infringement of copyright?

Could it be argued the Artist owns the copyright and merely used a machine to make a derivative work which Artist owns pursuant to 17 USC 106 et. seq.?

The issue is whether learning by scanning is in and of itself an infringement of copyright. We humans are entitled to view copyrighted works with our eyes and learn. Computers see by scanning. AI learns the techniques and can apply same to a picture a photographer has taken to create a work in the style of famous Artists. Is it a fair use to train a computer by simply allowing it to “view” all the works in digital format?

We are on the dawn of a new age.

Bicentennial Man walks among us.