How worried should we be - FB/IG scrapping for AI
-
This post is deleted! -
@Timbdsf said in How worried should we be - FB/IG scrapping for AI:
What concerns me most is when I heard FB and IG user posts will be scrapped to train AI. Is this actually true?
It’s true. If you're in Europe and want to opt out you should delete your account(s) before the June 26th deadline. US users weren't notified, possibly because US account holders have already been scraped. You can read more about the situation on Mashable:
https://mashable.com/article/meta-using-posts-train-ai-opt-out@Timbdsf said in How worried should we be - FB/IG scrapping for AI:
My husband also wonders that even if we remove content, it might not mean it's actually gone. Is that also possibly true?
Your husband may be referring to how servers constantly re-write files in order to ensure data is available. There's also file archives and backups of previous states for system recovery. Personal devices can save thumbnails of images for quick loading. So, yeah, he’s not wrong about the possibility.
Back in the early 2010’s, I warned friends not to use Twitter because people scraped images from it to slap designs on cheap cellphone cases and t-shirts. You can imagine how well that went over with others. If you were fine with the risks of living on the internet then, you should feel fine about the internet now. Assume your data was already mined.
If you want to keep your accounts, maybe consider just deleing your old posts. Start a new slate, posting images and information you wouldn’t mind too much if it were stolen. Like how brick-and-mortar stores display certain merchandise outside on the sidewalk.
-
I would delete your images, but not your account.
I recently saw an illustrator who was trying to get an account on Instagram deleted because someone was using her name and stealing her art. Don't relinquish your presence on these platforms, just control what's they have access to.
-
@willicreate there's an update on that: https://thehackernews.com/2024/06/meta-halts-ai-training-on-eu-user-data.html
regarding Cara or other alternatives (Bluesky etc), those are mostly for artists or maybe studios to look up portfolios but I don't think a lot of customers/"normal people" will create an account there to look at art.
So if the goal is to get sales through social media, afaik there's no real IG alternative (yet).
I'm not going to delete my account but I only used glazed (and nightshaded) images so far so if I'll ever post again there, I'm gonna do that further.
I'm also in the EU and we'll see how long Meta has to pause their AI training here and what comes out of the complaint situation. -
@Arkania Thanks for the article link. Glad to see a law acutally protecting citizens.
-
@willicreate Finally GDPR is good for something
-
@Arkania Bluesky doesn't have huge numbers of users (yet) but its certainly not a platform for artists/portfolios. It's actually a twitter alternative and has a growing kidlit community. I haven't had an account for long but i've already been followed by a few random non artists who just came across my posts. I hope Bluesky survives and grows because i think it's a great platform.
-
The bigger picture I see is there’s nothing stopping anyone else from posting our artwork on any platform that trains AI. And I don’t mean impostors. For instance, many people have shared my posts or posted their own photos of tshirts and books I’ve designed with my cartoon characters on them.
I also have a website with all my comics. I’m sure by now Google or any number of other platforms had already scrapped the heck out of it before I even heard of generative AI.
I really hate the unethical practices and self-entitlement of tech companies trying to make big bucks from scraping our art for their AI venture while pushing us into a corner.
-
@danielerossi I'm trying to figure out how to block Ai crawlers from my own website at least. That at least should be doable I think.
-
Hi, I'd like to confirm if my future Instagram posts will be safe from AI usage. I submitted a form to Meta through Instagram, stating that I do not want my images used for AI development since I am based in Europe. I received a quick response saying that my request is justified and will be followed in the future. Can I trust this, or should I ensure further protection by using Glaze or Nightshade?
-
@Mimi-Simon are you saying that whatever it being used to train the AI model is also going on personal websites and portfolios?
-
@Katt how did you build your website? Did you code it yourself or did you use a website builder platform?
-
@Katt Yes absolutely! I mean it's more likely to go on websites of famous artists but no reason not to be too carful. Heres what I found out to do:
create a new file in your websites files named "robots.txt" and put this text in it.User-agent: Google-Extended
Disallow: /User-agent: CCBot
Disallow: /User-agent: GPTBot
Disallow: /this will tell AI crawlers I don't want them there and they should go elsewhere. Some crawlers are designed to override this stuff but others are designed to obey. I included Chat GPT's crawler and Google ai crawler and a generic crawler. I couldn't find out the names of any others for example like Midjourney's crawler. If i find out more I'll add them.
-
Ok I did some more research and found a list people have been compiling of all ai crawlers they've been able to discover. feel free to use this list and start blocking away.
https://github.com/ai-robots-txt/ai.robots.txt/blob/main/robots.txt
-
This post is deleted! -
@Mimi-Simon I will keep that in mind, thank you
-
@thomas-young I used Wix, and bought a domain name to connect to it.
-
@Katt here is some information I found on protecting your Wix website:
Wix:
• SEO Tools: Wix provides built-in SEO tools where you can manage robots.txt and meta tags.
• CAPTCHA: Wix Forms includes options to add CAPTCHA to prevent automated submissions.
• Security: Wix includes built-in security features and monitoring to help protect your site from malicious bots.Wix:
1. robots.txt: • Go to the SEO settings in your site dashboard. • Customize the robots.txt file as needed. 2. Meta Tags: • In the SEO settings, add meta tags to the header section of your pages. 3. CAPTCHA: • When editing a form, go to the Form Settings. • Enable CAPTCHA to protect the form from spam. -
@thomas-young thank you! i will keep this in mind when I get my site back up. How do you all know this stuff? Please tell me you looked it up and didn't know it already
I tried to learn coding for about 2 seconds a couple years ago and was like "nope". lol.
