I keep seeing webmasters talking about how to block AI scrapers (through user agents and IP blocks) and not enough webmasters talking about the far better option of rigging their site to return complete gibberish or transgender werewolf erotica* when AI scrapers are detected.
*depending on which one you think is funnier to poison the AI models with
@foone i’ve been returning a 302 to a 10GB binary file from hetzner’s speedtest page, but honestly.. maybe I should?
all of my pages already contain prompt injection (in multiple places, even)
@domi @foone short term thinking! There's always going to be more scrapers who haven't learned their lesson.
If we can poison the well at scale, we can collapse the business model 
...
Who am I kidding?... I'm sure they'll have models selecting which data is "legit" and it will get better at detecting the "transgender werewolf erotica" over time... The stupidest of arms races
Alice Averlong🏳️⚧️
dmi 💽 
Emelia/Emi
LisPi
sunz*
Nic Roland 
Nathanif ($http_user_agent ~ 'GPTBot|ChatGPT\-User|Google\-Extended|CCBot|PerplexityBot|anthropic\-ai|Claude\-Web|ClaudeBot|Amazonbot|FacebookBot|Applebot\-Extended|semrush|barkrowler|PetalBot|meta-externalagent|meta-externalfetcher|facebookexternalhit|facebookcatalog') {
return 308 https://nbg1-speed.hetzner.com/10GB.bin;
}
if ($request_uri ~ 'wp-content|wp-login\.php|wp\-includes') {
return 308 https://nbg1-speed.hetzner.com/10GB.bin;
}
it’s a bit opinionated, the list of bots includes not only AI but also other BS. I include this into other files, inside server { }