iiinova — where your digital self keeps evolving

4 posts

Owen Boswarva @owenboswarva.bsky.social · 04/09/2026

Guidance on Using CC Licenses in an AI Ecosystem creativecommons.org/2026/09/03/g... from @creativecommons.bsky.social #IPlaw #techpolicy #genAI #opendata #openweb

The retreat toward restrictions has real costs. When people share content under more restrictive terms, the result is a more limited pool of publicly accessible resources (or commons) for all of us. This means less material is available for researchers, educators, archivists, and the public interest uses that depend on open access.

With that in mind, decisions around open sharing must not be taken lightly. Given that copyright exceptions often mean CC license restrictions don't apply to AI use in the first place, using a more restrictive license isn't effective in preventing it. In fact, a more restrictive license may prevent the exact types of human sharing and reuse licensors are hoping to allow, while at the same time putting no restrictions against AI use. As a result, we continue to guide communities to share as openly as they can, using CC licenses and tools to the extent that it is consistent with their values and goals.

010

Owen Boswarva @owenboswarva.bsky.social · 29/08/2026

UK celebs are called for new property rights to protect their voices from AI www.bbc.co.uk/news/article... Scottish singer Sandi Thom said she was signing the letter because her voice was "unique to me and was handed to me at birth". Not sure the science supports that but okay #genAI #IPlaw

Actors including Matt Lucas, Hugh Bonneville and Nicola Coughlan have written to the UK government demanding greater protection against their voices being used by AI.

The performers are asking Prime Minister Andy Burnham to introduce legislation which would give every person in the UK a legal right to own their voice.

Voice clones are created by training AI on recordings of real people. Quick versions can be made in minutes from a few seconds of audio, but high-quality copycats tend to require hours.

020

Owen Boswarva @owenboswarva.bsky.social · 27/08/2026

Awkward! When AI art has no author: Study finds generated images often can't be traced to training data news.mit.edu/2026/when-ai... Outputs of generative diffusion models are often unattributable doi.org/10.1038/s414... by Zheng Dai and David K. Gifford #genAI #IPlaw #attributiondecay

The scientists identified a phenomenon they call attribution decay, where the more data a generative model is trained on, the less any individual training example matters to any particular output. It feels counterintuitive, but at sufficiently large scales, they find, you can often remove any single image from the training data, or every image by a given artist, or every photograph of a given person, and the generated sample doesn't change.

And if removing something changes nothing, the researchers argue, it can't be said to be responsible for anything.

051

Owen Boswarva @owenboswarva.bsky.social · 24/08/2026

Anna's Archive Owes $340 Million, Lost Several Domains, but It's Still Online torrentfreak.com/annas-archiv... annas-archive.gl #IPlaw #webarchiving #shadowlibraries #openweb

Looking more broadly at the enforcement action that has taken place over the past months, we see that U.S. courts have run into their jurisdictional borders on the Internet.

This likely comes as a disappointment for rightsholders, but it also offers a clear takeaway.

U.S. courts can't reach domains registered beyond their jurisdiction. That's likely to increase calls for site-blocking legislation, a measure the industry has long favored and that remains high on the political agenda in the United States.

000