> I can get away with exposing a port on my router
using wireguard is awesome, but I wouldn't say deploying and maintaining it lacks complexity exactly. and you didn't get rid of attack surface, you just made it your responsibility ;)
Attack surface almost entirely goes away if you just require a good port knock first! I do that with my protection SSH servers and it works just fine. Also prevents anyone from knowing there even is an SSH server there. Or any server if it doesn't respond to pings either, and only uses reverse tunnels to accept public connections
With AI, the complexity doesn't really matter anymore. Codex or Claude can easily configure complex site-to-site setups and peer configurations on mobile devices. I did this by hand before and it was hard and difficult to get right. Now I just tell AI and it does so perfectly, checks and verifies things. AI made life so easy.
No amount of claude can handle the key thing that Tailscale fixes about wireguard, all of the connection to connection logic is handled be it direct, via all sorts of weird hole punching or via relays. Wireguard itself is horrible to do this with, half the time you're just looking at `wg status` wondering why the packet counter isn't incrementing today. It can never achieve the magic that is a direct connection between plane wifi and a host that's behind NAT on the ground.
This is true, but I still wouldn't discount the possibility of mistakes. It's good to know how your home network works. Even Opus 5.5 can make some pretty dumb mistakes, particularly if it doesn't have the full picture.
I would not personally expose my devices over a VPN that I self-hosted and set up purely through AI unless I felt like I understood how it worked completely. But that's probably just personal paranoia.
> Most of the Tailscale SSH issues we've identified concern privilege escalation. In some circumstances, someone already authorized to connect over Tailscale SSH may be able to perform actions beyond the permissions of their configured local user, including potentially gaining elevated privileges.
"Most"? "In some circumstances"?
Makes me think that they could also be investigating a full on unauthorized connection issue, or am I reading too much into it?
I think they're just writing cautiously because they're midway through the investigation and don't want to give false confidence. Doesn't seem worth reading too much into it until they provide a more thorough write up.
Obviously tailscale has some great devs, security is complex.
There is some weaselly language.
In some circumstances default config? with password? this is a condition leading to the next
someone already authorized to connect over Tailscale SSH they have access, but it may be limited or restricted, although scope could be beyond the intended
may be able to perform actions beyond the permissions If there is a full moon and your brother in law's second cousin is wearing a hat, he may be able to use your upstair's thermostat provided you have a second story.
A security researcher colleague of mine was/is exploring an unauthenticated avenue along this line. I don’t have any update, she told me she didn’t feel comfortable exposing information to anyone other than Tailscale staff until she’s confident in what she found or didn’t.
There are too many features, and lines of code (probably more than OpenVPN). I'm pretty sure there are all kinds of zero days, with this attack surface and feature set.
I use it extensively, and I would rather pay them a fee if they could review thr code.
Another possibility is as lock down mode, disabling fancy and risky features.
With the success of Tailscale I would like to be sure that their repositories and scenarios are continuously reviewed. I understand that there is always a next hop in cybersecurity though.
I like Tailascale as a product but I decided to move my homelab to plain WireGuard.
I wasn't using any of the additional features, I can get away with exposing a port on my router and I am changing devices rarely.
News like this make me happy I removed a lot of complexity and attack surface
> I can get away with exposing a port on my router
using wireguard is awesome, but I wouldn't say deploying and maintaining it lacks complexity exactly. and you didn't get rid of attack surface, you just made it your responsibility ;)
Attack surface almost entirely goes away if you just require a good port knock first! I do that with my protection SSH servers and it works just fine. Also prevents anyone from knowing there even is an SSH server there. Or any server if it doesn't respond to pings either, and only uses reverse tunnels to accept public connections
Port knocking is idiotic and amounts to Biff Tannen's "only I know how to start this car".
Is there a way to automate port knocks in your .ssh/config? I think that’s the main reason why I haven’t added it to my home server.
With AI, the complexity doesn't really matter anymore. Codex or Claude can easily configure complex site-to-site setups and peer configurations on mobile devices. I did this by hand before and it was hard and difficult to get right. Now I just tell AI and it does so perfectly, checks and verifies things. AI made life so easy.
No amount of claude can handle the key thing that Tailscale fixes about wireguard, all of the connection to connection logic is handled be it direct, via all sorts of weird hole punching or via relays. Wireguard itself is horrible to do this with, half the time you're just looking at `wg status` wondering why the packet counter isn't incrementing today. It can never achieve the magic that is a direct connection between plane wifi and a host that's behind NAT on the ground.
This is true, but I still wouldn't discount the possibility of mistakes. It's good to know how your home network works. Even Opus 5.5 can make some pretty dumb mistakes, particularly if it doesn't have the full picture.
I would not personally expose my devices over a VPN that I self-hosted and set up purely through AI unless I felt like I understood how it worked completely. But that's probably just personal paranoia.
From the sounds of what info they felt was safe to provide, this wouldn't have impacted you anyway.
Don’t think that refutes their point.
I think they're just writing cautiously because they're midway through the investigation and don't want to give false confidence. Doesn't seem worth reading too much into it until they provide a more thorough write up.
The underlying writeup can be found at https://tailworm.com/
10,000 words of claudish, it's not worth reading.
Obviously tailscale has some great devs, security is complex.
There is some weaselly language.
In some circumstances default config? with password? this is a condition leading to the next
someone already authorized to connect over Tailscale SSH they have access, but it may be limited or restricted, although scope could be beyond the intended
may be able to perform actions beyond the permissions If there is a full moon and your brother in law's second cousin is wearing a hat, he may be able to use your upstair's thermostat provided you have a second story.
A security researcher colleague of mine was/is exploring an unauthenticated avenue along this line. I don’t have any update, she told me she didn’t feel comfortable exposing information to anyone other than Tailscale staff until she’s confident in what she found or didn’t.
Not a tailscale user so I am a bit lost here: where does SSH fit into a Wireguard based VPN service?
As added candy to their daemon to get you into their ecosystem
https://tailscale.com/docs/features/tailscale-ssh
I self-host a tailnet with Headscale and could not be happier: https://headscale.net/stable/
Sure, that doesn't mean it's not vulnerable.
> Organizations using headscale v0.29.5 have no mechanism to disable the forwarding behavior at the center of the finding.
https://techjacksolutions.com/scc-vendor-rollup/tailscale-he...
There are too many features, and lines of code (probably more than OpenVPN). I'm pretty sure there are all kinds of zero days, with this attack surface and feature set.
I use it extensively, and I would rather pay them a fee if they could review thr code.
Another possibility is as lock down mode, disabling fancy and risky features.
With the success of Tailscale I would like to be sure that their repositories and scenarios are continuously reviewed. I understand that there is always a next hop in cybersecurity though.
I think it's so invasive that they ship this per default. But thats the business model i guess.
I've always had to manually enable this feature. I've never experienced it on by default across multiple platforms and deployments.
Do they ship this by default? I thought it had to be enabled: https://tailscale.com/docs/features/tailscale-ssh#configure-...
How do you mean? You have to run a command to enable tailscale ssh, it isn't on by default.
Tailscale SSH is not enabled by default.