A bipartisan group of senators introduced a new bill to make it easier to authenticate and detect artificial intelligence-generated content and protect journalists and artists from having their work gobbled up by AI models without their permission.
The Content Origin Protection and Integrity from Edited and Deepfaked Media Act (COPIED Act) would direct the National Institute of Standards and Technology (NIST) to create standards and guidelines that help prove the origin of content and detect synthetic content, like through watermarking. It also directs the agency to create security measures to prevent tampering and requires AI tools for creative or journalistic content to let users attach information about their origin and prohibit that information from being removed. Under the bill, such content also could not be used to train AI models.
Content owners, including broadcasters, artists, and newspapers, could sue companies they believe used their materials without permission or tampered with authentication markers. State attorneys general and the Federal Trade Commission could also enforce the bill, which its backers say prohibits anyone from “removing, disabling, or tampering with content provenance information” outside of an exception for some security research purposes.
(A copy of the bill is in he article, here is the important part imo:
Prohibits the use of “covered content” (digital representations of copyrighted works) with content provenance to either train an AI- /algorithm-based system or create synthetic content without the express, informed consent and adherence to the terms of use of such content, including compensation)
This is actually pretty cool for small artists, but how would it handle things like iFunny and such adding watermarks to shit they don’t own in the first place?
It’s certainly very bad for small artists. Can I ask why you think it would be good?
To answer why it is bad for small artists: Money for license fees will mainly go to major content owners like Getty, or Disney. Small artists will have to go through platforms like Shutterstock or Adobe, which will keep most of the fees. At the same time, AI tools like generative fill are becoming ever more important. Such licensing regimes make artists’ tools more expensive. Major corporations will be able to extract more value.
Look at Adobe. It has a reputation for abusing its monopoly against small artists, right? Yet Adobe pays license fees for images on its platform, that it used for training its AI tools. Adobe has also created a provenance standard, such as this bill wants to make mandatory. This bill would make Adobe’s preferred business model the mandatory standard by law.
Disney won’t sell licenses.
They’ll keep their monopoly on 95% of the training data on the market for entertainment content, so they can accelerate their workflows using those tools, and everyone else is now trying to compete with their giant wallet and their extra tooling you’re not allowed to compete with.
I am not understanding. They will now be even more effective at what they do and will be building a tech stack that no one else has access to and you think this is a win for the little guy?
If you owned a store, then a Walmart setup next to it and Walmart rolled out some new barcode system that your scanners didn’t work on how the hell would that benefit you?
Not only are the already bigger, they are getting efficiencies they aren’t sharing with your business. You want open protocols, you want tool sharing, you want the market as a whole to be getting more efficient. You don’t want some company that already won able to pulverize by sheer size anyone who competes.
No, it’s very obviously not a win for the little guy.
If you can’t learn from public performances, art dies. That’s all of art for thousands of years.
So what will happen to art if we only disallow AI models from learning from copyrighted performances, whilst still allowing them to do so from public domain and licensed works (and obviously not changing anything for humans who seek inspiration)?
Disney will own all of it.
Because they already own most of it and now have a massive productivity advantage.
Nah. They will cross licence with the other big players effectively closing the market to anyone they don’t bless.
They already own all the big players.
I don’t think that’s the kind of watermark being talked about here, Kol.
The National Institute of Standards and Technology would be called upon to, quoted from the COPIED ACT Summary, facilitate development of guidelines for voluntary, consensus-based standards and for detection of synthetic content, watermarking and content provenance information, including evaluation, testing and cybersecurity protection. I believe we’re talking about the unseen, math-y, certification and (I imagine) cryptography kind of digital watermark, not the crappy visual edits made by iFunny and co.
In fact, since it also says:
The content in question might reach e.g. iFunny already “signed” and they wouldn’t be able to remove that.
Of course, I’m saying this without actually fully understanding what fits under covered content (digital representations of copyrighted works). Does my OC on deviantart count as covered content? I think so, but I couldn’t tell you for certain. If anyone can help me understand this, please, that’d be really nice.
And finally, as was already said by others, I think this does nothing about all the crap companies already did to artists, since the law can’t affect them retroactively. It’s not that cool for small artists, since they’ll still be abused, except big tech would have the legal monopoly on abuse.
I mean no disrespect by this: did you read the article? I’m genuinely curious how you got the iFunny idea.
That is any digital representation of any copyrighted content. If it’s on DeviantArt, it is a digital representation. By creating something, you own the copyright. A notable exception is when you do the work for hire, in which case it probably belongs to your employer. (Or if the work lacks human creativity, in which case it is public domain.)
Anything on DeviantArt is almost certainly covered content, with the possible exception of AI generated images, or re-uploads of public domain content.
For reference: https://www.law.cornell.edu/uscode/text/17/102
Thank you for clarifying, and with a reference, too. That’s pretty much what I thought. It’s great to have confirmation, though.