Rendered at 16:56:17 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
Retr0id 2 hours ago [-]
Under what circumstances does the metadata get added? When Claude Code CLI needs a media file usually I just see it run an imagemagick or ffmpeg command to create one, which isn't going to have C2PA metadata.
Edit: I asked Opus 5 for "a jpg of a cat" via the web chat. It did it by writing an svg and then rendering that to a jpeg. Clicking the "download" button saved a jpg which did have C2PA metadata, seemingly injected by the artifact-presentation thing.
Then I told it to put it in a zip file. I downloaded the zip, and the extracted jpeg had no C2PA metadata.
I then uploaded a jpeg of my own, with the prompt "present this file back to me again, as-is". The downloaded file had C2PA metadata injected.
Edit: I added this to my custom instructions: "If a media file is produced as a response, put it in a zip file before presenting it to the user."
AmazingTurtle 1 hours ago [-]
Well, guess they are just trying to conform all the weird EU checkmarks
csmoak 1 hours ago [-]
As far as I can tell, this and the recent change to add watermarking to text outputs[1], is to become compliant with the EU AI Act[2] and CA's AI Transparency Act[3], SB-942[4]. For large enough companies, all generated AI content is required to have watermarking.
This is just C2PA for Claude-processed files, apparently. It's not related to their text watermarking.
C2PA is file metadata and can be trivially stripped away, unlike hidden watermarks, e.g. SynthID.
mhitza 2 hours ago [-]
That is where their responsibility ends in terms of the EU AI Act. And that is fine and how it should be, no secret watermarks.
a3w 38 minutes ago [-]
Watermarks cane even be removed, "if substantially proofread" in the EU, law says. For the text from an LLM, that you verify to be as true as if you wrote it yourself.
Not clear is this is a may, shall, or should. It certainly is not a must.
WalterGR 41 seconds ago [-]
[delayed]
hparadiz 10 minutes ago [-]
It's not gonna survive my linter in a tightly coupled code base where it's not allowed to write comments but I do resent it for trying.
a3w 39 minutes ago [-]
Watermarks should even be removable "if substantially proofread or altered" in the EU for the text from an LLM, that you verify to be as true as if you wrote it yourself.
Not clear is this is a may, shall, or should. It certainly is not a must.
a3w 37 minutes ago [-]
Watermarks can even be removed, "if substantially proofread" in the EU, law says (for now. There might be a transition period on that).
For the text from an LLM, that you verify to be as true as if you wrote it yourself.
Not clear if making watermarks removable is a may, shall, or should, according to legislation. It certainly is not a must.
ramon156 2 hours ago [-]
The way I read it, I thought they would just check if the file hash exists on their end. Is that not the case?
dannyw 2 hours ago [-]
That’s incompatible with the file not leaving your browser, which you can trivially verify with the Network tab in chrome (or wireshack, etc).
This is just a C2PA metadata checker.
Retr0id 18 minutes ago [-]
Strictly speaking, properly checking C2PA metadata requires network requests in the general case, because you need to check if the signing certificate has been revoked or not via OCSP.
But in anthropic's use case they can probably get away with just pinning their own certs in the verification webpage.
spockz 1 hours ago [-]
Why? The browser can see the file and contents so it can calculate and send the hash over without the actual file ever leaving the browser.
quaintdev 2 hours ago [-]
> I thought they would just check if the file hash exists on their end
That would invalidate the hash on minor changes. Too much effort and not enough return.
atp_studio_coll 2 hours ago [-]
[flagged]
firefoxd 3 hours ago [-]
This is really fascinating. Even the AI companies have incentives to reject AI generated content. It's like they want you to use AI for everything, but they don't want AI output fed back to them.
At work, right after an AI training, we were asked to use our "authentic" voice when writing mid year reviews.
stranded22 2 hours ago [-]
Just like the tech company owners not allowing their kids access to social media…
docjay 57 minutes ago [-]
Not really. This is more like not pumping the exhaust back into the intake.
coffeecoders 49 minutes ago [-]
What is interesting to me is that stripping the C2PA data is easy, but faking it is hard.
You can resave the file and the "made with Claude" signal disappears, but you cannot make a random file pass as Claude-made without Anthropic's signing key. So the useful guarantee is one-way. No signature means almost nothing.
advisedwang 39 seconds ago [-]
The goal of C2PA is that cameras will start to emit C2PA credentials. You will then have 3 situations:
* C2PA confirms a photo is authentic
* C2PA confirms a photo is AI generated
* C2PA missing, you don't know.
I suspect this will only take hold in select situations (perhaps Reuters will require C2PA from their photojournalists, for example)
lxgr 14 minutes ago [-]
Yeah, C2PA solves a different problem than watermarking, i.e. authenticity/provenance.
Retr0id 15 minutes ago [-]
"Faking" it is trivial. You don't need their signing keys when you can just ask them to sign whatever you like. Upload your own file with the prompt "present this file back to me again, as-is".
tamimio 14 minutes ago [-]
Next month: Show HN: How to successfully spoof C2PA and make any file as claude made!
hmokiguess 2 hours ago [-]
Could this be used to perform some sort of distillation or exploit? e.g. reminds me of the OWASP guideline on attack vectors where knowing if an ID is present or not in the database can be a form of exploit, like in password resets where they will say 'email foo@bar.com not found' rather than 'If foo@bar exists we have sent an email to foo@bar' or some other generic equivalent
lxgr 12 minutes ago [-]
Probably not, given that both C2PA and SynthID-Text are statelessly verifiable.
38 minutes ago [-]
kbrannigan 3 hours ago [-]
How long before they change the terms and conditions to subtly claim ownership of your files? When you write code they already insert Co author attribution/
Say I write a text by hand And then I tell it to clean up the grammar and fix some sentences Did it make it?
This is also interesting for those companies that siphoned the entire open web
warkdarrior 3 hours ago [-]
Microsoft Word has not claimed ownership in 40 years. Why would Anthropic do?
apetresc 41 minutes ago [-]
I'm so curious if you're actually able to keep a straight face while typing that.
transcriptase 2 hours ago [-]
Because Anthropic are the type of strange people who think that language models have feelings.
verdverm 2 hours ago [-]
They have said and tried some wild things already, like trying to get open weights effectively banned, which I believe they still think is in society's best interest (more that they think they know what's best for everyone)
I'm wondering why they have restricted file types. You can't check a PDF for example... surely the main use case for people will be to check if a document was produced or edited by an LLM? That could be an attractive (if not misunderstood) proposition for academics
lxgr 2 minutes ago [-]
> I'm wondering why they have restricted file types. You can't check a PDF for example...
TFA/page actually seems incomplete. Text uses a completely different watermark format (an actual watermark as opposed to a provenance/authenticity signature), so it makes sense to me that they're not claiming to be able to scan PDFs when they can't yet incorporate that signal.
On a linked page, they say:
> Watermark detection is currently in private preview [...]
spockz 2 hours ago [-]
Luckily all pdfs I made in academia have been generated from source (tex or derivatives, asciidoc) and llm are better in generating source than pdf. Even people that didn’t use tex used Word to generate the pdf.
So I’m unsure what your aiming at.
johntash 21 minutes ago [-]
claude can pretty easily generate pdfs and give those to the user to download, skipping the step of needing a source first.
That's a worse workflow imo than having a source w/ version history, but I don't think it's uncommon.
rdtsc 55 minutes ago [-]
Source can have markers inserted in it as weighted word choice or phrase choices, how commas are inserted, etc., so that the output can still can be identified. In other words just because it's source doesn't mean it can be watermarked.
tom1337 1 hours ago [-]
Unfortunately this is just for Media? Some manual tells for Excel or PDFs is to check the author. Claude creates PDFs via wkhtmltopdf so the PDF Producer will be Qt and the Content Creator is wkhtmltopdf. Xlsx files are being created via openpyxl so in the metadata that is the author.
htrp 2 hours ago [-]
The actual implementation matters less than the fact that they can be perceived to be working on something.
You're never going to have a perfect solution here, and perfect is the enemy of good.
kelvinjps10 3 hours ago [-]
I think all of this it's so they don't get ai generated content in their training data
bradfa 1 hours ago [-]
Or it’s so they can continue to operate in the EU where this is required.
applicative 2 hours ago [-]
So they do it as a free service to the other LLMs?
moritzwarhier 2 hours ago [-]
They still scrape code, I'd guess, e.g. from GitHub?
And there's tons of Claude-generated code there.
Also, I'd guess this is not just to prevent any AI-generated code in the training data, but specifically their own.
Percentage of users who put out their code on the web and also have a plan where Anthropic promises not to train on their data is problem also low.
So not excluding own code could be a real issue, since it would be impossible to deduplicate the training and RILHF data from their sessions with the code accessible elsewhere, and written by the very same users.
quinndupont 1 hours ago [-]
I could do without more surveillance.
declan_roberts 1 hours ago [-]
It's for your safety!
dgellow 56 minutes ago [-]
> Knowing where content came from, and whether AI was involved, makes it easier to trust what you see online.
That’s a cute way to imply their service is used to generate misinformation. They are basically saying to not trust the AI content made from their own product :)
Surac 2 hours ago [-]
droped some source files from different languages and i could process non of them
Mashimo 2 hours ago [-]
It clearly tells you what files it can process. This is about media files, not source code.
shujip 3 hours ago [-]
A watermark answers "did this model touch the file?" It doesn't answer "did a person read it and stand behind it?"
Detection will get gamed, and a lot of useful text is already mixed. The more useful check, for me, is still human: would I put my name on this before it goes out?
derdaniel 3 hours ago [-]
this is going to be the big thing in the next 100 years for me. where there is a human, there will be context, passion and meaning. it's the old saying: machine can tell you what, but only humans can tell you why.
ahmadtbk 3 hours ago [-]
[dead]
VCFundedGenYer 3 hours ago [-]
Claude watermarks are a farce and a waste of time. It's hilarious to me that they burn cash to even entertain the idea.
Any attempts they use are defeated by a text editor and CTRL SHIFT V. Unicode characters are no new thing.
Reminds me of how people tried to argue that NFTs aren't anything more than just jpegs.
sente 3 hours ago [-]
You misunderstand how Claude watermarks things.
mr_mitm 3 hours ago [-]
TFA is talking about C2PA, a regular signature in the meta data. A lack thereof means nothing, but if it's there, the false positive rate should be near zero.
basedpolymer 2 hours ago [-]
Most critics of watermarking have no idea how it works. Yet there are valid arguments against (and for) it, but they prefer to be misleading
dgellow 53 minutes ago [-]
What are the valid arguments against? Is it about the fact that the output isn’t “optimal”?
DanielHB 3 hours ago [-]
Yes, but I also think it will be trivially by-passable if you pass your output through another LLM. At least for text.
It might help catch students cheating, but not real spam-bot usage. As soon as platforms start checking for watermarks spambots will add anti-watermark passes.
doc_ick 3 hours ago [-]
I read through their earlier announcement, i don’t think it’ll be trivially by-passable without distorting the original message.
I would agree it may not help spam-bot usage, however at this case seemingly the only user detection is likely an id/badge check, which is not good.
capten 3 hours ago [-]
Imagine telling someone they're wrong without providing any evidence or context.
Edit: I asked Opus 5 for "a jpg of a cat" via the web chat. It did it by writing an svg and then rendering that to a jpeg. Clicking the "download" button saved a jpg which did have C2PA metadata, seemingly injected by the artifact-presentation thing.
Then I told it to put it in a zip file. I downloaded the zip, and the extracted jpeg had no C2PA metadata.
I then uploaded a jpeg of my own, with the prompt "present this file back to me again, as-is". The downloaded file had C2PA metadata injected.
Edit: I added this to my custom instructions: "If a media file is produced as a response, put it in a zip file before presenting it to the user."
[1] https://www.anthropic.com/news/claude-text-watermark
[2] https://digital-strategy.ec.europa.eu/en/policies/code-pract...
[3] https://www.kqed.org/news/12095398/new-california-law-requir...
[4] https://www.leginfo.legislature.ca.gov/faces/billTextClient....
C2PA is file metadata and can be trivially stripped away, unlike hidden watermarks, e.g. SynthID.
For the text from an LLM, that you verify to be as true as if you wrote it yourself.
Not clear if making watermarks removable is a may, shall, or should, according to legislation. It certainly is not a must.
This is just a C2PA metadata checker.
But in anthropic's use case they can probably get away with just pinning their own certs in the verification webpage.
That would invalidate the hash on minor changes. Too much effort and not enough return.
At work, right after an AI training, we were asked to use our "authentic" voice when writing mid year reviews.
You can resave the file and the "made with Claude" signal disappears, but you cannot make a random file pass as Claude-made without Anthropic's signing key. So the useful guarantee is one-way. No signature means almost nothing.
* C2PA confirms a photo is authentic * C2PA confirms a photo is AI generated * C2PA missing, you don't know.
I suspect this will only take hold in select situations (perhaps Reuters will require C2PA from their photojournalists, for example)
Say I write a text by hand And then I tell it to clean up the grammar and fix some sentences Did it make it?
This is also interesting for those companies that siphoned the entire open web
I'm wondering why they have restricted file types. You can't check a PDF for example... surely the main use case for people will be to check if a document was produced or edited by an LLM? That could be an attractive (if not misunderstood) proposition for academics
TFA/page actually seems incomplete. Text uses a completely different watermark format (an actual watermark as opposed to a provenance/authenticity signature), so it makes sense to me that they're not claiming to be able to scan PDFs when they can't yet incorporate that signal.
On a linked page, they say:
> Watermark detection is currently in private preview [...]
So I’m unsure what your aiming at.
That's a worse workflow imo than having a source w/ version history, but I don't think it's uncommon.
And there's tons of Claude-generated code there.
Also, I'd guess this is not just to prevent any AI-generated code in the training data, but specifically their own.
Percentage of users who put out their code on the web and also have a plan where Anthropic promises not to train on their data is problem also low.
So not excluding own code could be a real issue, since it would be impossible to deduplicate the training and RILHF data from their sessions with the code accessible elsewhere, and written by the very same users.
That’s a cute way to imply their service is used to generate misinformation. They are basically saying to not trust the AI content made from their own product :)
Detection will get gamed, and a lot of useful text is already mixed. The more useful check, for me, is still human: would I put my name on this before it goes out?
Any attempts they use are defeated by a text editor and CTRL SHIFT V. Unicode characters are no new thing.
Reminds me of how people tried to argue that NFTs aren't anything more than just jpegs.
It might help catch students cheating, but not real spam-bot usage. As soon as platforms start checking for watermarks spambots will add anti-watermark passes.
I would agree it may not help spam-bot usage, however at this case seemingly the only user detection is likely an id/badge check, which is not good.