Google To Defend Generative AI Users From Copyright Claims (reuters.com) 16
Google said on Thursday that it will defend users of generative AI systems in its Google Cloud and Workspace platforms if they are accused of intellectual property violations, joining Microsoft, Adobe and other companies that have made similar pledges. From a report: Major technology companies like Google have been investing heavily in generative AI and racing to incorporate it into their products. Prominent writers, illustrators and other copyright owners have said in several lawsuits that both the use of their work to train the AI systems and the content the systems create violate their rights. "To our knowledge, Google is the first in the industry to offer a comprehensive, two-pronged approach to indemnity" that specifically covers both types of claims, a company spokesperson said. Google said its new policy applies to software, including its Vertex AI development platform and Duet AI system, which generates text and images in Google Workspace and Cloud programs.
Whats the strategy here for Google? (Score:1)
Re: (Score:3)
Make money, lock people into ecosystem to make more money in the future.
That said, I think Google is going the wrong way here, I think Apple is going to destroy them on this front too as a result. Rather than indemnifying customers, they should train only on licensed and public domain content. This is what Adobe is doing and will almost certainly be what Apple will do, because Apple doesn't want to tarnish their image. Now Google is setting themselves up to be the evil and legally risky alternative. Indemni
Re: (Score:3)
Naruto v. Slater, and the current behavior of the copyright office denying copyright to AI generated works and Google's legal win in being able to scrape the web for their search engine being a use of copyrighted materials that does not require the copyright holders permission gives a pretty strong foundation for Google's belief that there is minimal risk.
Additionally, If Google wants to go full evil, they will use their AI to find the works that influenced the art that the person is claiming the generated
Re: (Score:2)
That does nothing to protect against (statutory) damages for all the copy steps during training. The many copy steps leading up to and during training are unprotected by any of the existing DMCA exemptions, precedent from lawsuits about search or implicit license doctrine ... it's almost certainly going to be judged plain infringing (and statutory damages alone will be crippling, no need to prove actual damage).
The resulting outputs are likely not to be considered infringing, but they will then by fruits fr
Re: (Score:3)
The people suing are not only claiming the generated works violate their copyright, but that their copyright is violated by AI reading their works.
If they successfully make that claim, they will kill the current LLMs and make generative AI much harder.
So, it is clearly in the interest of Google, Microsoft, and others to support the targets of these lawsuits.
For the record, I hope Google and Microsoft prevail. Reading and learning are not "copying" any more than a human "copies" an image onto their retina wh
Re: (Score:2)
Re: (Score:2)
Microsoft Copilot vacuuming up the source code to Linux, then spitting it out on demand?
That's not the way it works. Copilot can't spit out a kernel.
Without copyright, F/OSS cannot exist
Without copyright, we don't need F/OSS. The point of the GPL is to fight copyright with copyright.
default copyright license applies ("all rights reserved") and you're restricted in what you can do with the code.
No. The output from AI can't be copyrighted. So anyone can do whatever they want with it.
then I can feed Linux in, get a AI version of Linux out
You are vastly overestimating the capability of LLMs.
So is AI infringing or not?
The output may be. But training is not.
If I sell copies of your book, I'm infringing. If read your book, I am not. It should work the same if the reader is a computer.
Re: (Score:2)
If you read my book, someone should have paid for the copy you are reading, be it you directly, or a library. It's the paying that grants you the right to make copies on your retina.
Re: (Score:2)
If you read my book, someone should have paid for the copy you are reading
Except they aren't reading books. They're reading websites that any human can read for free.
Also, if you steal a book and read it, that is stealing, not copyright violation.
Re: (Score:2)
I know. I was commenting only on your analogy about books.
Re: (Score:2)
These are only my opnions not being a lawer
> I mean, imagine Oracle sending the whole of Oracle Linux through AI, to get Oracle Linux AI, a completely "original" reinterpretation of F/OSS code by AI, and without the pesky requirement to give away source or other things.
You mean like Phoenix Technologies did for the IBM bios?
https://www.computerworld.com/... [computerworld.com]
https://en.wikipedia.org/wiki/... [wikipedia.org]
>So is AI infringing or not?
AI can read and learn from the code like a human but if produces code that is
Soo, another "pledge"... (Score:2)
We all know these are utterly worthless.
Obvious motive ... (Score:3, Interesting)
The strategy for Google and others is obvious, bypassing copyright through lawyering up. Let's be real, few artists and creatives have the deep pockets to go toe-to-toe in a copyright fight against these companies, hence they're bypassing the legal system altogether and making sure their investments in AI have a chance to return a tidy profit.
Re: (Score:2)
And very few ai users have deep pockets to go in a copyright fight against the big media companies.
Wild, WILD antitrust violation. (Score:3)
Google at it again (Score:3)
Ah, the company once bought YouTube and made all YouTubers presumed violators of copyright when a complaint is made until proven innocent through a Kafkaesque process is now extending the convention to everyone who hasn't signed the YouTube terms of service too.
If an artist doesn't want to sign agreements with allowing Google to use their work and doesn't want their work to be taken and thrown into the Google Content Making Machine maybe they shouldn't have let others post it on online, falsely claiming it as their own for a quick buck?
That way they wouldn't be on the opposite end of a lawsuit with a company that's wealthier than several countries combined, right?
This is wrong in so many ways I could spit but the way IP law works it's a dismally foreseeable ambition for a large internet company.