Let Residents 'flag' bots that aren't registered as Scripted Agents
Beatrice Voxel
I get a lot of bots just dropping onto my land at random. It's a residential parcel on an adult island region, so I value my privacy. Of the bots I've encountered since I got the parcel, all of them have had blank profiles and randomized names, with nothing that indicates they are a bot. They show up, they stay put for ~10 seconds, and then they port out again. I'm not running a shop, or a club, or any kind of venue for random people to visit. This is my home in SL.
Rather than clog my Casperlet ban list with bot accounts that likely will be scrapped and recreated, it would be nice if when I'm looking at their profiles, there was an action "Report Suspected Bot". This would allow residents to notify LL as to WHICH scripted accounts aren't actually set as such, and also provide a mechanism for auto-banning bots that aren't welcome (via the Scripted Agent designation).
Proposal for the actual report:
Date/time the report was made, by whom, and where (SLURL). Add a small comment section for the resident to say "this is what happened, this is why I'm reporting this account as a bot."
Note that such a report would not flag the account as a bot directly, only indicate that it is suspected of being such. Multiple reports by different accounts at different locations would of course weight the 'score' - this would prevent a single person from maliciously reporting people as bots.
Log In
Kim Primdashian
This is absolutely needed! There users who make alt accounts using popular bot names: ie. BonnieBot7472826, just to spy/have access to others lands and parcels for whatever reason.
Zy Butcher
I am upvoting this not because I want this feature, but because bots are problematic and annoying on the long run. Especially if they are not flagged as scripted agents.
Skyler Pancake
There's 100% absolutely no need for this.
LL can easily figure out what accounts are scripted agents on their own. They don't need residents to report them. Lou Netizen has already created a framework for this, one which could be replicated and improved by the lab quite simply.
The truth of the matter is that LL does not care about bots and see's no reason to police them. It's the same reason we don't have parcel control over allowing scripted agents even though that was promised by Patch two years ago now. They came out with a half-baked solution by giving region control, rewrote their policy page, and then ignored the situation until the public outcry died down. It's how the team post Ebbe operated and still does.
Also, for ducks sake, stop using ban orbs to "ban" bots. Anyone who bought an anti-bot ban orb was scammed. Ban orbs rely on scanning who's present before to then determine if the person should even be kicked. So either your constantly lagging your region by scanning every millisecond or the bots are doing what they want within the few seconds before they get kicked anyways. Just use the parcel or region ban list, as that ACTUALLY prevents them from TPing to your location at all.
Beatrice Voxel
Skyler Pancake Sadly, my parcel ban list is full. So is the region's ban list. So I have to rely on your standard Casperlet security orb. It reports on any avatar's arrival and gives them (and me) 30 seconds to obtain a guest auth or leave, at which point if they're not supposed to be there they get sent home. I also get an IM letting me know that So And So has just arrived, and a dialog to add them to guests, temporarily allow the visit, or ban them outright.
My workflow with these (yes it happens often enough for there to be a workflow!) is to check their profiles - random all-lowercase names with NOTHING in the profiles, no feed, no picks, no RL data, get a note attached - when did they they appear, if their profile had any info, and if they get added to Casperlet's parcel ban list. This is how I catch repeats - the next time they show, I see my note on their profile's notes page. At that point, they get banned and I note the date that happened.
Has this slowed down the flow of bots? No, it hasn't. I still get repeat visits. I ban 3-4 bot accounts per week.
In addition, I have a 1 square meter subdivision of my parcel (not policed by the orb), in the extreme corner of the sim where tping to coordinates 0,0,0 will land you if there's no LP. I've put an LP above this point, in a tiny little box called the "Oubliette" - there's nothing in it but a transporter pad locked to my use only, and a visitor recorder. This catches anyone who was ejected from land on the region, but didn't have a home point defined. I get ~10 of these per week as well - most are brand new accounts with no information, likely yeeted from my main parcel, or one of the other parcels.
As I understand, BonnieBots ident themselves as such - surprisingly I haven't seen any of those. Most of what show up appear to be mass-generated accounts with random names, purpose unknown.
I've had ONE AI companion bot show up, it IM'd me as soon as it arrived with some kind of "Hi, I'm your AI assistant" prompt. It got ejected and banned with no reply.
As you've pointed out, I'm not really preventing these accounts from arriving. I am at least able to log them.
Gwyneth Llewelyn
I definitely support this idea, even if — technically speaking, that is — Peter Stindberg is absolutely correct and there is close to zero probability of 'catching' a rogue bot quicker that way, compared to the already-existing mechanisms that are in-place. In other words: it's very rare and unlikely that a rogue bot will do a 'visit return', so to speak; it will be banned by LL before that happens!
Nevertheless, I still support this idea, and the reason is simple: it gives residents a
sense of control
over their environment. It's irrelevant if this is actually useless or not; it gives residents the feeling that they're not absolutely powerless when facing hordes of 'bots, but that they can actively contribute to keep the number of rogue 'bots to a minimum. In other words: it seems to empower residents to become pro-active in the 'bot wars' instead of being mere helpless victims.None of the above is true, of course, but it's a cosy, warm feeling that one gets when clicking on a flag somewhere and getting the notion that you're contributing towards your well-being and that of all the other
human
residents.Gwyneth Llewelyn
If all else fails, Beatrice Voxel, nothing like taking a peek at Monkey Canning's Zeus anti-griefer, anti-alt, and anti-robot tool. There, I've just done some free advertising! 😄 Note that Monkey's tool (running in-world since 2016) is self-contained and does not allow 'fake flagging' or 'revenge flagging' or gaming the system in any way; and it also doesn't capture any of the user's data that are forbidden under the ToS (such as digital fingerprinting via IP address, etc.). It's essentially a heuristic algorithm that takes into account one's avatar
behaviour
, according to some sets of metrics (fine-tuned over the years), to immediately detect suspicious accounts — and inform all participants in the Zeus network of their presence, a bit like how email anti-spam filters work iRL.That might be a valid alternative in your case — just set things up to whitelist your close friends (and/or partner) and check everybody else, kicking them out if they aren't validated by the system according to a set of rules that
you
define. Since those actions are logged on Zeus' back office webpage, you can then, at your leisure, fill in as many abuse reports you like, with the confidence that you'll have the almost exact time of the occurrence on record. Or
you can simply ignore all the logging and let the system work automatically on your behalf.It's cool. Perfect? No, no system is perfect. But it might complement LL's own rogue 'bot detector tools and at least keep them out of your way until LL deals with them.
RestrainedRaptor Resident
Gwyneth Llewelyn I took a look at the Zeus site, but it doesn't even explain how to implement it. Surely they should be pointing to a marketplace profile or something.
Gwyneth Llewelyn
RestrainedRaptor Resident I totally agree that the instructions fall a bit short of the mark; I was also unimpressed with that Drupal installation with a standard template from the early 2010s, and an utter lack of organisation... but alas, that's what we get.
Anyway, the entry for the item (or, rather,
system
— it has several items for different use-cases) is definitely publicly available on the SL Marketplace, but it takes a bit of searching that should be absolutely unnecessary when a plain, simple link on the home page would suffice.RestrainedRaptor Resident
Gwyneth Llewelyn Thanks for the link. I tried searching for 'MonkeyCorp' but I guess I needed to search for 'Monkey Corp'!
AlettaMondragon Resident
As Nya mentioned replying to Peter, this "flagging" is already available via an Abuse Report:
Maybe a "Bot / Scripted Agent Policy Violation" category should really be added to the Report Abuse feature so that it would be clear to us all where to report such issues, especially since in recent years the bot infestation became overwhelming.
Personally I have no problem with bots whose operators are transparent enough and publish their research, but many bots are impossible to identify unless it's your area of expertise to do so. This caused quite the panic among many people and it should be dealt with because now some resident-made devices kick real people from places too, not caring about collateral damage just so they can keep some of the bots out. A parcel-level bot access restriction option and a limit on how many bots a bot operator can have should be implemented.
Peter Stindberg
I am researching bots in SL for about 4 years now. My observation is that rogue bots (those who are confirmed bots, but have the scripted-agent flag not set), have a very short lifespan, sometimes only a day, before they get deleted/retired by the Lab or their operators.
After about a year since the scripted-agent status was introduced, my observation was that all "known" bots had the flag properly set. The rogue bots - therefore one has to assume - are rogue on purpose. Their operators WANT them to violate the policy, and know only a high churn rate - create them fast, let them run for a day or two - then retire them, is how they can achieve that.
There is nothing we residents can do, and little the Lab can actually do. They could make signups harder (and we all don't want that), they can enter an arm's race (and we see with spammers/antispammers how well that works), or they can state an example and litigate one bot operator into oblivion as a deterrent to others.
No reporting-mechaism will help, as charming as the idea is.
N
Nya Jules
Peter Stindberg
I agree 100% and would like to add that LL encourages residents to write an Abuse Report if someone suspects an account to be a non declared scripted agent, therefore "flagging" an account is already possible. I don't like the idea of automated bans (no matter if it's about bots or any other alleged TOS violations) at all because this would get abused and it's also not reliable at all.
Skyler Pancake
Peter Stindberg Need to disagree due to technical clarification. ROAMING bots not marked as scripted agents are likely to be caught. Non-roaming bots not marked can easily go unnoticed. Other than that, 100%. If you'd ever like to chat with the Bonnie group about bots, hit us up. We're always down for a fun conversation.
Peter Stindberg
Skyler Pancake I fish non-marked
roaming
bots out of my detector grid on an almost daily basis. They are short-lived, exist for a day or two, then get deleted. Whether they get deleted by their own operators, or by the Lab, is a technicality. The sheer fact that two years after Bonniegate the situation is worse than before. Sadly, I have to concur with your assessment in an earlier comment: The Lab has no real interest in doing something about it. They could use the large-scale statistical analysis that Lou Netizen or Darling Brody do (which is prone to false positives), they could use the micro-detector-approach I take (which is prone to false negatives), they have access to other data we do not have access to (login-ip and fingerprints come to mind) and could curb it, combined with a "show trial" against one of the more egregious operators as a warning to others.
I always said there is nothing the Lab can do other than making signups harder (which we all don't want). This position has shifted recently. Maybe it needs a second Bonniegate, that forces the Lab's hand again. But every time the Lab's hand gets forced, the underlying problem did not change, only legitimate use cases got more difficult.
And this, Skyler, is the one huge grudge I hold against the Bonnies: That they became so brazen, so annoying, so smug, that they forced the Lab's hand, and at the same time normalized having armies of bots swarming the grid. On a sidenote: It also costs you credibility. Your remark somewhere in this thread, that people should stop using bot-banning-orbs - something I wholeheartedly agree with and which is something I debated hotly with Darling Brody about - it turns into the opposite when it comes from someone affiliated with a bot-group, because then I think 'Well if THEY are against it, maybe it works after all'.
I am sad for all the useful, benevolent, interesting, helpful roaming bots that gave up in the wake of Bonniegate. I knew one bot operator who even had his RL address in the profile of the bots, for transparency. He ran his 5 Cookiemonster bots for quite a while, but shut them down in the wake of Bonniegate because harassment against his person and anit-bot-orbs increased.
Once a problem becomes so pressing, that policy needs to be made against it, we all lose. That's what the Bonnies achieved. Well done.