The news this week has a lot of stories about AI. For those of you that are fed up with reading one frustrating headline after another, this one is interactive, and for all of you who have a strong opinion on exactly the kind of issues we’ve been hearing about for months now. The US Copyright Office currently has a consultation open on AI and copyright. The deadline for written feedback is October 18th.
Submit your views to the US Copyright AI Consultation
The scope of the consultation includes the use of copyright material in training datasets. It also includes how copyright should apply to works that have been generated using AI. Interestingly, it also covers an area that’s new to the legislative framework: the imitation of style. This will apply to voices and performances as well as artwork and words.
This isn’t about misrepresentation. Rather, it’s about what to do with works generated by prompts that ask, for example, for an artwork created “in the style of” a particular artist. This of course is also tied in with the use of materials in training sets.
This is very timely given Amazon’s new policy on AI, which allows AI-generated work provided the use of AI is declared. The protection afforded to any AI-generated work you upload will depend on the outcome of this consultation.
But people who use AI at present should be under no illusion. A high-profile case this week shows that at present you cannot protect such work. Matthew Allen’s Theatre D’opera Spatiale made history when it won an award at last year’s Colorado State Fair. But the Copyright Office has just ruled that it can’t be registered because the artist used the generative tool Midjourney. At present, that means AI which has been developed from scraping artists’ content is itself totally fair game for scraping.
If you have opinions on any of these issues, I would encourage you to take part in the consultation.
Always stay up to date with self-publishing news with the weekly ALLi news podcast.