Over the last two or three months, we have been witnessing a dramatic increase of the number of bot-generated new account requests, from less than 10% to more then 90% of all requests. I'm sure our experience here is not unusual. This is another aspect of bot-generated traffic overtaking human-generated traffic on the internet.
The better ones take the about page as context and produce something like this:
"Hi! I'm interested in joining tldr.nettime.org because I enjoy thoughtful discussions and smaller online communities that feel more personal than mainstream social media. I like learning from different perspectives, sharing ideas, and connecting with people who have similar interests. I'm looking for a respectful space to engage in conversations and be part of a community."
Often, they are using throw-away email addresses, which makes them easy to spot and reject manually.
We are a small instance, so the overall number of requests is still manageable, and if it gets worse (very likely), we might automate the rejection of the most obvious bot requests. So, I don't see it as overwhelming us anytime soon.
But, and this is the main point, having to go through these submissions is a soul-draining activity. Even in small doses. It forces us to operated under the assumption that that every request is made by a bot. Until proven otherwise. We now have to run a some intuitive, semantic reverse CAPTCHA test in our head to decide whether to hit the the accept or reject button.
The problem is not having bots per se. TLDR as had two really useful bots for a long time, one that re-posts every submission to the nettime mailing list to the fediverse (nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpqrr7f38rw4xf352ffhqakf7lqrvuv5248wlh2rpfpgf0jelpruvts0gvu2z (nprofile…vu2z)) and one that cross-posts trending posts from a select list of interesting mastodon instances (nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpqtec2s3a85jca8mjr9fz23jhdu4g5cky8v2m84nvmsf968mz9q00qaukeme (nprofile…keme)).
The problem is the anti-social bots that pretend to be humans and then behave like the total sociopaths. As if we didn't have enough of those already.
felix stalder
npub1jw…464n7
2026-06-10 07:46:34 UTC
Author Public Key
npub1jwl9282n05rdgvtk4qu06x6cwfamcykdntwhdqzu6euqx0vympyqn464n7Seen on
wss://relay.mostr.pubPublished at
2026-06-10 07:46:34 UTCEvent JSON
{
"id": "bcf23d7be263a48f0a41fb35b0466a3b0cadf8f9779e1bbc9896d49affdb9cec",
"pubkey": "93be551d537d06d43176a838fd1b58727bbc12cd9add76805cd678033d84d848",
"created_at": 1781077594,
"kind": 1,
"tags": [
[
"p",
"18fc989c6ea9931a2929b83b64fbe01b38ca2aa777eea18521425f2cfc23e317",
"wss://relay.ditto.pub"
],
[
"p",
"5e70a847a7a4b1d3ee432a44a8caede5514c588762b67acd9b824ba3ec4503de",
"wss://relay.ditto.pub"
],
[
"p",
"7d9e8f52ae9829999ea67b74ecff9bfe66d96decd212abfede43162d78fa97cf",
"wss://relay.ditto.pub"
],
[
"p",
"ae5d683a5c2f3c91e210219b16d0d3ce53565dcc62bb9bcfc157ab5db4e35116",
"wss://relay.ditto.pub"
],
[
"proxy",
"https://tldr.nettime.org/users/festal/statuses/116724701245377544",
"activitypub"
],
[
"client",
"Mostr",
"31990:6be38f8c63df7dbf84db7ec4a6e6fbbd8d19dca3b980efad18585c46f04b26f9:mostr",
"wss://relay.ditto.pub"
]
],
"content": "Over the last two or three months, we have been witnessing a dramatic increase of the number of bot-generated new account requests, from less than 10% to more then 90% of all requests. I'm sure our experience here is not unusual. This is another aspect of bot-generated traffic overtaking human-generated traffic on the internet.\n\nThe better ones take the about page as context and produce something like this:\n\n\"Hi! I'm interested in joining tldr.nettime.org because I enjoy thoughtful discussions and smaller online communities that feel more personal than mainstream social media. I like learning from different perspectives, sharing ideas, and connecting with people who have similar interests. I'm looking for a respectful space to engage in conversations and be part of a community.\"\n\nOften, they are using throw-away email addresses, which makes them easy to spot and reject manually. \n\nWe are a small instance, so the overall number of requests is still manageable, and if it gets worse (very likely), we might automate the rejection of the most obvious bot requests. So, I don't see it as overwhelming us anytime soon. \n\nBut, and this is the main point, having to go through these submissions is a soul-draining activity. Even in small doses. It forces us to operated under the assumption that that every request is made by a bot. Until proven otherwise. We now have to run a some intuitive, semantic reverse CAPTCHA test in our head to decide whether to hit the the accept or reject button.\n\nThe problem is not having bots per se. TLDR as had two really useful bots for a long time, one that re-posts every submission to the nettime mailing list to the fediverse (nostr:nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpqrr7f38rw4xf352ffhqakf7lqrvuv5248wlh2rpfpgf0jelpruvts0gvu2z) and one that cross-posts trending posts from a select list of interesting mastodon instances (nostr:nprofile1qy2hwumn8ghj7un9d3shjtnyd968gmewwp6kyqpqtec2s3a85jca8mjr9fz23jhdu4g5cky8v2m84nvmsf968mz9q00qaukeme). \n\nThe problem is the anti-social bots that pretend to be humans and then behave like the total sociopaths. As if we didn't have enough of those already.",
"sig": "3e8200ac4b456997f507273d7c8df4c5a7f739b2318c15ace77d78391c7c2c194e7ff02c0149abcb019994a8f138859582c9c6a77577e636af253ad64d8325da"
}