Is this forums admins doing what they can to protect artists to have their content not scraped by AI developers?
I think this is something that is 100% an issue and I would like to know what level of security if any is being employed to prevent anyone here being stolen from… especially considering you could download anyone’s piece by clicking on the image and downloading original…
You can’t really protect an image that is shown on a website from bots that search the web for images. Even if you put visible watermarks over the images, you wouldn’t stop these bots and make their job unfeasible, at least not as long as such a visible watermark made the view virtually impossible for us humans. This is because these crawlers are designed for pattern recognition, it would also be easy for them to calculate recurring watermarks from images and could subsequently “enjoy” these images more unclouded than we humans, who are the actual target group, could. A download option is not even necessary for these crawling bots, if they can identify an image on a website, they can also “view” this image, they would only benefit from downloads in resolutions above the resolution of the website.
Website operators have few options. You could allow downloads only to logged-in members, perhaps develop features that don’t display an image until you click and hold a button, or even require certain buttons to be clicked in the correct order, but doing so can also alienate your users.
If you want to protect your images from such scrapers, then paint them in your head and don’t describe them to anyone when they first see the light of day (or the dark of night), then the images can also be photographed by some unthinking person and subsequently end up in unauthorized hands. However, the emphasis is on CAN!
The only thing that is certain is that nothing is certain once it has left your body.
If one wants to steal your intellectual property, you can’t do much against it. If you want to show your pics to others, you are at risk.
Michelist
Sadly we can’t do much for web scraping. Html is like an open book we can’t hide things nor we can make it closed. Having watermark is also futile as these AI companies can remove it for example AI output had Shutterstock or getty watermark but AI companies have now deviced method to remove that. If a person wants to steal they can steal there are bots which can screenshot and take pictures too.
There are blogs that do not allow the download of images… however, there is a way around this using the browser’s own resources.
If i remove my account here will the images be deleted as well?
Yes if you wish we can remove everything and after 4 days everything will be gone from the backups too. I will however suggest you don’t do it since it will be sad to see your artwork removed from here. you can remove your artworks and keep the account or you can remove everything and close the account. But It is up to you as you own your artwork and not us you only grant us the right to show it to people via this forum.
On a general note you can’t keep your images off the AI companies anywhere on the open internet unless you lock it behind a login and even then still people will login and save your images to the drive. Unless you personally verify each person you can’t be sure who will save your images and feed it to the AI.
Or good old screenshoting
Yea i mean thats all fine and good, but there are places like art station that are allowing an opt out function… how that gets implemented IDK but its at least considerable… it might be smoke and mirrors…
but even as an artist that is growing… I do not wish that my progress is stripped away by people who want to capitalize on a software that rids my personal ability to make a living… this forum or others… makes no difference to me…
Hi
The difference between a classic portal and artstation website:
Classic portal
You can put a rule in robot.txt to inform AI scrapper to not use image for training
Up to the AI scrapper to ignore or not the rule
ArtStation, DeviantArt, and other big place
They sell hosted pictures for AI training and/or use hosted pictures for their own AI implementation training unless you explicitly set an opt-out option on your artworks
Concerning external AI Scrapper, maybe as they want to sell the artworks they might put robot.txt and firewall/proxy rules to try to block some AI Scrapper but as it’s not really difficult to create a scrapper that is able to pass everything, it’s something useless for whom really want to use their huge artwork database
In synthesis: whatever the place you put your artwork on the web, as soon as it’s on a public access (you can access it from a browser) you can see your artwork being used for AI training
Grum999
@raghukamath maybe add CCBot in exclusion list
Won’t block all AI Scrapper but at least, seems Stable Diffusion used this one…
Grum999
Option Out and robots.txt protect as well as the I am over 18 years old button that is supposed to protect XXX websites from being accessed by minors, namely not at all if someone wants to steal from you.
Michelist
okay well i want to remove my content then…
You have to tag @raghukamath for that.
Michelist
Unfortunately you have to remove it from everywhere on the internet if you want to be 100% sure. But I can understand your decision, made the same on DeviantArt.
Sure 
Done!
Our user list page are already blacklisted from bots by default. Still I think the bots can crawl the posts whichw e can’t deny since this is a forum and search indexing is good for a forum. Now which bot is search and which is AI that is hard to distinguish. if you have any other disallow rules like this just ping me. We can do whatever we can to discourage this madness.
Should I delete your artwork posts and keep any help related one?
Also I don’t know how that NOAI tag is working if I know I will enable it on this website unless it means editing the forum software. in which case we need to make a feature request to the Discourse forum software
In that case, you can not publish your work anywhere online, because those who want to disobey rules like robots.txt, or option out declarations, or TAG’s like NoAI, those can not be stopped by this, they will disesteem these declarations of will.
You would trust in their good will, fair-mindedness, respectability, and as we all know by now, they have more than proven their dishonorability!
And in case you choose an offline gallery to display your work, you have to make sure that no visitor has anything at hand like smartphones or an old-fashioned camera. And thinking of smart glasses equipped with camera and always online, the nightmare for any paranoid artist comes true, you can’t display anything anymore.
Michelist
Your concern is justified and completely understandable: no one wants to be ripped off. However, I remembered something a drawing teacher told me: if you create a character, say a witch and someone comes along today and steals that drawing… what happens?
You think you’ll never be able to create anything better than that character ever again. But that doesn’t happen. Tomorrow you will be able to draw a better witch than that one. Because you will have evolved and will be able to produce something better.
Hi
I just took a look on what Artstation have implemented for that.
If user manually add NoAI predefined tag, then in HTML page, the <header> markup will contain the following meta markup: <meta content="NoAI, index" name="robots"> to indicate to scraper the artwork should not be scrapped 
Not sure how this can be interpreted by scrappers, not even sure if it’s a standard…
Also, their robots.txt is empty, no rule is provided to try to inform bots to not download content.
If tag is not added, then I suppose:
I didn’t had so much artwork on artstation (but deletion took me a long time), but for all people with hundred or more artworks, it’s probably just hell to update them one by one…
Grum999
If it was my scraper I would just aim right for the img html-tags, everything else is overhead except maybe the artwork tags when you’re doing it for AI.
