To start this, here’s the evidence I have:

103.4.251.127 - - [30/Oct/2025:09:23:41 +0100] "GET /ai/possible-bot.html HTTP/1.1" 200 20542 "-" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/116.0.0.0 Safari/537.36"
103.196.9.229 - - [30/Oct/2025:20:23:33 +0100] "GET /ai/possible-bot.html HTTP/1.1" 200 29374 "-" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/116.0.0.0 Safari/537.36"
3.210.114.189 - - [09/Nov/2025:06:08:32 +0100] "GET /ai/possible-bot.html HTTP/1.1" 200 3675 "-" "Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; Amazonbot/0.1; +https://developer.amazon.com/support/amazonbot) Chrome/119.0.6045.214 Safari/537.36"
217.113.194.108 - - [09/Nov/2025:15:25:56 +0100] "GET /ai/possible-bot.html HTTP/1.1" 200 957 "-" "Mozilla/5.0 (compatible; Barkrowler/0.9; +https://babbar.tech/crawler)"
2a03:2880:f800:2:: - - [08/Nov/2025:05:30:47 +0100] "GET /ai/possible-bot.html HTTP/1.1" 200 28594 "-" "meta-externalagent/1.1 (+https://developers.facebook.com/docs/sharing/webmasters/crawler)"
44.195.145.102 - - [02/Nov/2025:02:42:13 +0100] "GET /ai/possible-bot.html HTTP/1.1" 200 4126 "-" "Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; Amazonbot/0.1; +https://developer.amazon.com/support/amazonbot) Chrome/119.0.6045.214 Safari/537.36"
74.7.227.6 - - [01/Nov/2025:11:44:43 +0100] "GET /ai/possible-bot.html HTTP/1.1" 200 428 "-" "Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; GPTBot/1.2; +https://openai.com/gptbot)"
34.196.237.236 - - [29/Oct/2025:08:24:19 +0100] "GET /ai/possible-bot.html HTTP/1.1" 200 4890 "-" "Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; Amazonbot/0.1; +https://developer.amazon.com/support/amazonbot) Chrome/119.0.6045.214 Safari/537.36"
114.119.140.71 - - [30/Oct/2025:00:29:25 +0100] "GET /ai/possible-bot.html HTTP/1.1" 200 37764 "https://www.bleen.dev/page/2/" "Mozilla/5.0 (Linux; Android 7.0;) AppleWebKit/537.36 (KHTML, like Gecko) Mobile Safari/537.36 (compatible; PetalBot;+https://webmaster.petalsearch.com/site/petalbot)"
114.119.128.57 - - [04/Nov/2025:06:35:34 +0100] "GET /ai/possible-bot.html HTTP/1.1" 200 13377 "https://www.bleen.dev/posts/using-chatgpt-to-cheat-on-stack-overflow" "Mozilla/5.0 (Linux; Android 7.0;) AppleWebKit/537.36 (KHTML, like Gecko) Mobile Safari/537.36 (compatible; PetalBot;+https://webmaster.petalsearch.com/site/petalbot)"
114.119.153.208 - - [04/Nov/2025:17:28:55 +0100] "GET /ai/possible-bot.html HTTP/1.1" 200 16376 "https://www.bleen.dev/tags/selfhosting" "Mozilla/5.0 (Linux; Android 7.0;) AppleWebKit/537.36 (KHTML, like Gecko) Mobile Safari/537.36 (compatible; PetalBot;+https://webmaster.petalsearch.com/site/petalbot)"

Now, where does this data come from (apart from my Apache server log)? On my other website (you can see it in the referer section of the log) I have an “invisible” div with a link to a “page”, which is explicitly forbidden for anybody via robots.txt. Which these bots clearly ignore. Now, the fancy thing with this “page” is that it is actually the output of go-pot, in the author’s own words “A service for giving away secrets to bots… Probably slightly too many”. In addition to HTML, as here, it also gives you too much info in json-, SQL-, XML- and a heap of other formats, and I really recommend it for what I use it for, namely poisoning these malicious bots.

If I explicitly forbid your bot, crawler, or whatever you want to call it to go somewhere, I expect that to be followed. AI companies like OpenAI or Facebook (you didn’t think Facebook was in the business of social media, right?) don’t even give half a f**k about our privacy, our data, and our knowledge and will (ab)use it to feed their ever-hungry LLMs.

I can’t wait for the AI bubble to burst. It’s using my time, my server resources, and I bet I get some other negative effects from it as well now that they’re being used in everything from healthcare via insurance to banking. At least with media we usually can detect quite good when an AI has “written” an article or faked a photograph. Although that gap seems to be closing fast, and some people (usually stupid people, or far-right people, but I repeat myself) will believe anything these so called intelligencies tell them, that’s why their “leaders” are so happy to use them at every opportunity. But at the end, they’re just really good at guessing the next word, and if they can’t make a qualified guess, they pull something out of their artificial ass. With go-pot, I help them a bit with the latter part making it seem like it’s the former.