Below is an image of me in 2005 for the Chronicle of Higher Education’s coverage of my dissertation. Interestingly, the photographer told me that he was paid for the print version but not the digital one. This is odd since print images are so much more expensive than digital ones.

In 2005, I defended one of the country’s first media-rich, digital dissertations. My main reason for its digital nature had to do with visual literacy and fair use as free speech. Firstly, I argued that the digital is as amenable to images as it is to words (or, at the time, nearly so). But also, if we cite words but pay for or otherwise ‘clear’ images, I maintained, we effectively allow Hollywood to dictate who may speak and, perhaps more importantly, who is silenced. Since that time, I have made and taught remix video under the Doctrine of Fair Use and soon, the Audio Visual Lexicon project authors are producing a video of me speaking about copyright and fair use.

Unfortunately, however, when it comes to Generative AI, BIg Tech has been winning lawsuits brought about by content creators whose creative output is being used by LLMs (large language models) as training data for chatbots such as Chat GPT, Claude, Gemini et cetera, on fair use grounds. This is simply wrong-headed and results from a misunderstanding of the way LLMs work. This may be changing soon as Disney and Universal launched a suit against MidJourny, the text-to-image platform in June 2025 showing clear copyright infringement.

I am currently working on a manuscript that details the issues involved and will elaborate more here soon.