AI Training Opt Out

Known tags and settings suggested to opt out of having your content used for AI training.

robots.txt A copy-and-paste collection of tags to add to your own robots.txt. (You can automate generation of this file with darkvisitors.com)
meta-tags.html A copy-and-paste collection of tags to add to your own <head>
headers.txt HTTP headers you can add to your responses. This is more more involved and installation is outside the scope of this document.
ai.txt An alternative to robots.txt created by Spawning, the company behind haveibeentrained.com.
ip-ranges.txt Known IP ranges for AI crawlers. These will change over time, so links to the canonical source is included.
tdmrep.json A Web protocol, capable of expressing the reservation of rights relative to text & data mining (TDM)

Other Opt-Outs

OpenAI (Includes ChaGPT and DALL·E): You can opt-out of having your input and output to their services used to train by emailing your organization ID to support@openai.com. Note: This doesn't include any data they scraped to train their model.
StabilityAI: Stable Diffusion 3 will honor opt-out requests on haveibeentrained.com.
AWS: "AWS may be using your data to train its AI models, and you may have unwittingly consented to it. Prepare to jump through a series of complex hoops to stop it." -- How to Stop Feeding AWS’s AI With Your Data
Substack "If you do NOT want your publication to be used to train AI, open your publication, go to Settings > Publication details and switch it on."
Wordpress and Tumblr are both opt-out for your post content.
The Stack Find your repo(s) on Am I in The Stack? and then click Opt-Out at the bottom to open a request.