AI produced art and their copyright

OpenAI released its text-to-image generation model based on transformers architecture called DALL·E and this has generated a numerous amount of concerns and questions directly linked to copyright issues.

DALL·E uses a process called “diffusion”, which starts with a pattern of random dots and gradually alters that pattern towards a final output. DALL·E “trained” on approximately 650 million image-text pairs scraped from the internet, learning from that dataset the relationships between images and the words used to describe them but the company admitted that the system can sometimes create works that include trademarked logos or characters.

Now that we’ve established what DALL·E is and how it works, we must now turn our attention to what constitutes copyright protection.

For something to qualify for copyright protection, it must cover 3 parameters:

1. an original work of authorship;
2. fixed in a tangible medium;
3. that has a minimal amount of creativity.

This is to say that, technically speaking, AI generated works of art cannot qualify for copyright protection, but this also raises the question as to who the actual owner since ownership may not be clearly distinguished.

If we take the time to look into DALL·E’s own terms of use, we can note the following:

• “Your Content. You may provide input to the Services (“Input”), and receive output generated and returned by the Services based on the Input (“Output”). Input and Output are collectively “Content.” As between the parties and to the extent permitted by applicable law, you own all Input, and subject to your compliance with these Terms, OpenAI hereby assigns to you all its right, title, and interest in and to Output…”
• “Similarity of Content. Due to the nature of machine learning, Output may not be unique across users and the Services may generate the same or similar output for OpenAI or a third party…Responses that are requested by and generated for other users are not considered your Content.”

 This means that OpenAI assigns the right, title, and interest in and to the final image to the creator, provided the terms have been complied with. Copyright assignment permits a third party to take ownership of the copyright from the owner, or assignor. This means that the creator, the person who inputs the text prompt, is the owner of the image and has the rights to the final image. The creator becomes the owner of the images, regardless of whether those images have copyright or not.

OpenAI precociously bypasses most copyright questions through contract and their carefully and cleverly worded terms of use, making rather evasive references to intellectual property ownership

In the end, humans are the ones that make the final decision to use art generated by a machine (to illustrate their articles for example); therefore, AI clearly cannot grant permission for use of the work or hold a copyright for the same.

Any disputes arising from computer generated works by AI will need to be assessed on a case-by-case basis and will likely depend on the level of contribution of each of the parties in the creation of the work.