Roaming Bots and Privacy
What About My Privacy?!
Second Life is meant to be an interactive social environment, so it isn't very private about avatars.
Anyone operating a bot in Second Life is required to log into the account on the SL website and set its "Scripted Agent" status to true, indicating the avatar is run by software and not a human. As of June 2023, this property is directly detectable by inworld scripts, so, if you like, you can make (or buy) a security device that instantly ejects scripted agents out of your region/parcel.
In early 2023 Linden Lab updated its bot policy to indicate it "reserves the right" to shut down any bot operation it believes is violating its bot policy—which would presumably include failing to register as a scripted agent—or for violating "applicable privacy regulations." Linden Lab added the capability for private region owners to exclude all scripted agents, unless they're specifically allowed. If you believe a bot is violating these policies, file an abuse report (use the "Disturbing the Peace" category).
Otherwise, bots operate under the same privacy conditions that apply to "real" avatars. Bots cannot read your IMs (unless you IM them!), nor can they see your inventory or your friends lists. Bots can receive local chat within a range just like any other avatar. It would be unusual, but bots could record public voice (again, within a range). It would be highly unusual, but a bot could take screenshots or record video, subject to the same rendering limitations of a normal Second Life viewer. (Most bots run on non-graphical, low-resource clients that can't "see" anything, not tricked-out gamer rigs with nitrogen cooling and enormous GPUs, but there are exceptions.) Just like any other nearby avatar, bots can identify your active group. If you list groups or picks in your profile, bots can see that just like anyone else—same with your public "feed." Bots can use Second Life search features.
Bots can go anywhere that's publicly accessible, so that means the best way to keep bots off your land is to ban them. If you have a private estate, you can forbid all scripted agents. They won't be able to enter. (Roaming bots almost always teleport: only a handful walk or fly around.) If you see an unexplained avatar turning up, feel free put it on your ban list—your land, your rules! However, bans won't necessarily prevent bots from detecting your presence on a region: just like a "real" avatar, they can be aware of anyone on the same region (and, to a lesser extent, adjacent regions), regardless of whatever parcel(s) you or the bot are in. This is just how Second Life works.
Be aware that maintaining a banlist will require constant effort. Some estate owners and land businesses have attempted to automate this process with wildly varying degrees of accuracy. Some bot cohorts retire old avatars, replacing them with new members, or cycle through large numbers of avatars with only a fraction being active at any given time. If you're deeply concerned, consider using your land's access lists to permit only certain known-friendly avatars or group members to enter. You could also consider restricting access to avatars with payment info on file or barring avatars younger than 30 days, but those techniques won't be particularly effective and may keep out legitimate visitors. Some roaming bots have payment info on file/used, and most cohorts that cycle in new bots seem to plan more than 30 days in advance.
Why Don't You Publish a List?!
I don't publish a list of bot avatar names or keys because I don't need the drama, but it's also not my place. I am not responsible for setting or enforcing policies in Second Life. Also, if I published such a list, ten minutes later someone would copy it and offer an anti-bot tool on Marketplace for L$5000. As I see it, exploitive opportunists can do their own R&D.
If you think a bot (or any other avatar) is violating Second Life TOS, file an abuse report with your evidence. (Use the "Disturbing the Peace" category.) That's what I do. From mid-2017 through mid-2025, I've reported 51 semi-roaming bots for activities I could conclusively demonstrate were violating LL terms of service, and about another 140-ish non-roaming bots and bot-like accounts primarily used by griefers and scammers. (I should note that I'm not in shopping etc groups, so I almost never see grift bots.) This is a tiny (tiny) (tiny!) fraction of the bots in SL. There are (many) (many!) thousands of other bots associated with some inworld games I'd consider "persistant nuisances" that I'd quite like to report, but they've been behaving openly for years, have been reported by many people innumerable times, so it sure seems LL is fine with them. I'll not waste more of my time.
It is worth noting that Linden Lab has demonstrably done nothing about most bot issues I've reported. "Demonstrably" because the bots in question continue to violate LL terms of service, community standards, and/or bot policy.
Don't like the bot situation in Second Life? I can only suggest taking it up with the Lindens. (Yelling at me won't help, I promise.) As of mid-2022 an official recommended path is to file a support ticket with Linden Lab. Another might be to turn up at an appropriate "office hour". In mid-2026 Linden Lab set up a "Trust & Safety" meeting on the third Tuesday of each month, conducted on voice. (A previous "governance" office hour was discontinued; I don't know the details.) The Trust & Safety team won't discuss specific cases but seem willing to listen regarding policy. I doubt the Lindens feel current bot policies are set in stone, and would consider amending/updating them further based on community feedback. Linden Lab also maintains a Canny-based Feedback portal; users there sometimes try to gin up support for various bot-related ideas. Perhaps writing those up (or voting in favour of them) catches the Lindens' eyes once in a while.