Thought Shrapnel

As Stephen Downes notes, Creative Commons (CC) has announced a new framework for signalling preferences around AI training. Building on an IETF draft standard, the idea is that creators can use machine-readable tags to express whether their work should be used for training AI, and if so, under what conditions.

Echoing the existing CC licenses, the ‘conditions’ might be providing credit to others, a financial contribution, or open-sourcing the resulting AI models. This all sounds well-intentioned, and you’d think I’d be in support of it. I’m a CC Fellow, after all. But instead, it feels out of touch with the reality of how large AI companies operate.

Stephen has already pointed out that this approach creates a pretend layer of control as CC Signals are not legally binding — and those with the most to gain from ignoring them are the least likely to pay attention. Our collective experience with [robots.txt]9en.wikipedia.org/wiki/Robo…) should probably serve as a warning for this: respect for voluntary signals only lasts as long as it suits the interests of those scraping the content.

Creative Commons licenses were created a couple of decades ago to allow creators to share their work openly and freely. CC Signals is attempting to protect creators, which is admirable, but makes the relationship transactional and so shifts CC away from its roots in open sharing. As the framework lacks any real mechanism for enforcement, there’s little to stop powerful actors from simply disregarding these preferences. It’s shutting the barn door after the horse has bolted.

TL;DR: technical standards are useful, but without legal backing or industry buy-in, this is little more than a polite request. I think the commons deserves more than a new version of robots.txt that can be ignored at scale.

Since the inception of CC, there have been two sides to the licenses. There’s the legal side, which describes in explicit and legally sound terms, what rights are granted for a particular item. But, equally there’s the social side, which is communicated when someone applies the CC icons. The icon acts as identification, a badge, a symbol that we are in this together, and that’s why we are sharing. Whether it’s scientific research, educational materials, or poetry, when it’s marked with a CC license it’s also accompanied by a social agreement which is anchored in reciprocity. This is for all of us.

[…]

Reciprocity in the age of AI means fostering a mutually beneficial relationship between creators/data stewards and AI model builders. For AI model builders who disproportionately benefit from the commons, reciprocity is a way of giving back to the commons that is community and context specific.

(And in case it wasn’t already clear, this piece isn’t about policy or laws, but about centering people).

Source: Creative Commons