AI & Copyright

John Laudun
2 min readSep 7, 2023

Heather Froelich pointed out that the U.S. Copyright office is seeking comments on artificial intelligence and copyright. Here is the comment I left, and a link to do the same is below. The comment begins somewhat awkwardly because the Copyright Office wants to know if you are representing yourself or a third party, and I want to be clear that I was speaking for myself:

Photo by John Laudun. (No rights reserved.)

I am writing for myself, but as a folklorist I am also writing with profound respect, and sadness, for our national tradition of enabling private profit at the cost of the public commonwealth. Like the pharmaceutical industry raiding traditions around the world in order to develop better, perhaps life-changing, medicines, we have allowed the large language models behind most of the more prominent AI platforms to harvest knowledge of a lot of individuals without the individuals themselves receiving any acknowledgment, compensation, or share in the profit. Whether we call it “folk” or “mass,” we dis-enfranchise those who actually produce the materials from which we derive products.

We cannot fall back on user agreements which, in order for the basics of the web to work, had individuals consent to broad grants of copyright. We must acknowledge that most users posted texts, images, and other media assets to various platforms and sites in the interest of creating and maintaining various communities. That they were willing to be sold to advertizers, because that is the basis for American media production, should not in any way affect our consideration that their materials, and thus the people themselves to some degree, can simply be given to AI platforms. At least the social media platforms gave them something of value in exchange. AI platforms are already monetized, seeking rent for creating an abstraction of a city built of neighborhoods built by others.

We cannot know what will be the eventual outcome of the development of these AI platforms, and I don’t think referening the hype or the fear-mongering does any good here. What we can know is that a system’s integrity must be clear and checked throughout the process. Right now, we can say for certain that these systems were built without integrity when it comes to their data acquisition. If we do not figure this out, if we do not create useful guidelines for clarity and integrity, than we are somewhat dooming these systems to have further negative impacts.

If you would like to contribute to the conversation, the commenting period is open until October 18 of this year:

Yours in the struggle.



John Laudun

Cultural Informatics Researcher focused on Stories, People, Networks