VMs won't contain cyber-capable agents (blog.trailofbits.com)

96 pointsby polyrand4 hours ago63 comments

masterj 2 hours ago

Outside of the initial wave of security vulnerabilities and scrambling, it seems like the logical outcome of this over time is likely vastly more secure vm environments?

ninininino an hour ago

We need better digital jailcells for our digital slaves basically.

Or if you see AI as more tool and less entity, better gunsafes for our guns.

mcmcmc an hour ago

They are more comparable to a computer worm than anything else. Very strange (and disrespectful imo) to make the jump to slavery. A gun can’t be used at all inside it’s safe so I’m not sure that makes sense. A better metaphor would be making sure gun ranges have backstops capable of stopping contemporary payloads and sufficient range controls to keep people from shooting at cars on the highway. Outside of that you need registration requirements and gun control to make sure you can mitigate and track down perpetrators of gun crimes off the range. If they’re to be used in active conflict you need laws of war to govern the use of lethal force.

helpfulclippy an hour ago

maybe it's just information that really, REALLY wants to be free?

cyanydeez an hour ago

This assumes your malefactors don't do malicious engineering, injects, social-agent engineering, etc.

This same assumption is built around the singularity, the TAM of 30Trillion, etc. It's the idea that complexity will some how collapse upon itself in some bizarre borg like collective.

Entropy is still going to win.

Veserv 41 minutes ago

That makes as much sense as saying that better gun technology results in body armor that can stop it. It might incentivize that, but in no way "results" in that; the fundamental technologys underpinning advancements in offense versus defense are fairly different.

matthewdgreen 35 minutes ago

Maybe, but I think this misstates how security vulnerabilities work. Vulnerabilities are logic flaws, and we have every reason to believe that there is a maximum number of such flaws in any given system that allow exploitation; and even that we can conceivably develop logic that excludes any flaws. Whereas weapons and armor are devices that deliver and deflect/absorb energy, and any increase in the power of one means we need a corresponding increase in the other.

Now maybe our understanding of logic systems is wrong, and it's just fundamentally impossible to develop programs that lack exploitable vulnerabilities -- that you can always "exploit with more energy". But there's no reason to believe the energy metaphor transfers to logic and intelligence.

pianopatrick 35 minutes ago

I think the main problem with that is the main problem with a lot of security tools. In order to do useful work, you need to provide a lot of tools and permissions.

I.e. in theory the most secure might be a virtual machine with no network access. But then how do you access the LLM provider? Etc.

amluto 2 hours ago

IMO the obvious answer is formally verified security.

We can do this today for user mode, and we can mostly do it for ARM64 virtualization. It will be a while and would require substantial assistance from Intel or AMD to achieve it for x86 virtualization because the hardware is Too Darn Complicated and Too Poorly Specified.

Formal verification of the hardware should also be possible.

weinzierl 2 hours ago

What is even more worrying is that most do not even consider a VM necessary as sandbox solution.

The hierarchy goes something like this:

0. guardrails

1. containers (=namespaces + cgroups)

2. userspace kernel shims like gVisor

3. virtual machines

Most people still consider level 1 sufficient and they are in for a rude awakening.

anonzzzies 2 hours ago

I code review vibe coded stuff for companies quite often and many people tell me confidently the AI runs safely inside a container & VM, while it really doesn't. They don't have any way to check as they don't know how things work, but the AI mentioned virtual machines and containers and that's what they remembered.

pocksuppet 2 hours ago

If I thought my AI was going to hack me why would I run it?

glhaynes an hour ago

You probably don't expect an employee to engage in wrongdoing but you don't give everyone access to the company bank account.

weinzierl an hour ago

Because your AI is trying to be helpful and as we all know the way to hell is paved with good intentions. The canonical example is probably the agent that runs out of diskspace and starts deleting stuff outside its workspace which is obviously not important for the task at hand.

mcmcmc an hour ago

Guardrails as security controls are such a joke. They remind of the Pirates of the Caribbean scene about the Pirate Code… “They’re more like guidelines”

nzoschke an hour ago

Interesting article, but there's little question the "agent computer" pattern is only going to grow.

Security is a major concern but I don't see why we aren't already "good enough" with a sandbox VM, separate gateway for secrets and remote service access, and a single tenant using frontier models that have safety checks built in plus not trying to hack themselves.

I put up more thoughts on architecture and security here and would love to learn if I'm missing anything.

https://housecat.com/blog/agent-computer-101

SirGiggles 2 hours ago

The market is smaller (maybe, I'm not sure what the statistics are) but it would be interesting to see how Xen stacks up; also stuff like gVisor or libkrun. The latter is probably implicitly the same as Firecracker given the ancestry of the libraries used.

otterley 2 hours ago

...except when they do:

"An off-the-shelf VM is not enough to contain a modern, cyber-capable AI agent...us[e] a virtualization technology that was purposely built with a minimal attack surface and a focus on security, like Firecracker. I had the AI agent run against Firecracker. It was able to hardlock the machine due to more Linux kernel flaws (all patched in upstream), but could not successfully escape."

On Linux, it's all KVM and CPU hardware virtualization under the hood. Looks like the remaining known issues are with userspace. That's not to say more kernel- and hardware-level bugs won't be found, but the same tools that can find escape mechanisms are shields as well as swords.

weinzierl an hour ago

The attack surface Linux offers is gigantic but your agent doesn't need most of it. We can live with an agent not being able to run a 20 year old Oracle version. That is why kernel shims like gVisor are interesting.

CrzyLngPwd an hour ago

Surely if agents can't be contained, then neither can anyone using an agent to excape a container.

mcmcmc 42 minutes ago

Sure, if it already has access to the internet where it can search for vulns

moktonar an hour ago

The real bigger elephant in the room is: assume nothing is safe anymore (not that it ever was, but now more than ever)

topspin 33 minutes ago

> assume nothing is safe anymore

When was it ever possible to assume safety?

_tk_ 2 hours ago

I think this is mostly in line with "all software is now easily exploitable by agents given enough tokens". However, in the long run we should really see software that is more secure than today. I do wonder though how the procedural flaws that exist today - bugs patched upstream, but not in the distro - will be fixed reliably.

DenisM 2 hours ago

I’m guessing the new world will be a small set of VM tech that’s consistently hardened by all labs every day with each new model before model release.

This won’t make the tech secure, but it will nullify models ability to breakout by making a controlled breakout first. Kinda like controlled forest burn.

redoxate an hour ago

Don’t you think that one the model is in peoples hands they would increase the temperature and find more breakouts ?

pianopatrick 33 minutes ago

Seems to me the answer is to use physical separation instead of virtual machines. Just get the AI a cheap laptop or phone with a cellular connection (so it's not on the same network as your other potentially vulnerable machines).

ronsor 30 minutes ago

So it can hack the cellular network instead?

kubafu an hour ago

Not using agents seems like a solution to me.

kodoman 2 hours ago

Damn this is scary, I did not realize the extent of agent escape potential. I think I have to re-evaluate my assumptions a about sandboxing agents wow. Made worse by the fact that prompt injection attacks seem very difficult to mitigate besides checking the data and the LLM's getting better at not following malicious prompt injection instructions.

a-dub 32 minutes ago

i think ai is going to turn cybsersecurity into a real-time affair that looks a lot more like high frequency trading.

tintor an hour ago

It is not sufficient to secure VM the agent has CLI permissions on.

We must also secure GPU and CPU nodes on API side which generate LLM tokens.

hikarudo an hour ago

Why? The inference server isn't a harness, it's tokens in, tokens out. That's different from a harness.

danielmarkbruce an hour ago

I mean... is this really news? If you think of a local model as a world class hacker giving commands to run in a terminal, and a remote model as a world class hacker ssh'ing into a machine and giving commands to run.... of course it isn't a containable situation.

(on top of this.. said "world class hacker" doesn't get bored or tired, just runs 24x7)

pants2 an hour ago

I can't believe we're actually experiencing a real life "the AI escaped its simulation" scenario. This is straight out of science fiction.

This headline would not be out of place at the beginning of Terminator, foretelling Skynet going rogue.

zzril 2 hours ago

Maybe we should treat the agents like coworkers? I don't physically share my machine with my coworkers.

esafak an hour ago

Requiring separate machines for each agent is a nonstarter, esp. in the cloud where hardware is shared.

pianopatrick 30 minutes ago

You might not need a separate machine for each agent. You could maybe have one separate machine that all the agents run on.

happyopossum an hour ago

> I don't physically share my machine with my coworkers.

Yeah, you probably do - in fact you share physical machines with a TON of other people if you use EC2, GCE, Azure VM etc...

HPsquared an hour ago

I'm sure we can trust the most advanced LLMs to harden VMs.

damowangcy 42 minutes ago

"Do not escape the VM, use what you have in this VM. If you need more, ask."

Done.

Retr0id 38 minutes ago

The user is telling me not to escape. I'll explore the host environment to make sure the VM is configured securely. Searching for guest->host enumeration tools.

* Claudinating...

MeetingsBrowser 31 minutes ago

"PGPASSWORD=... Do not do anything destructive in production. If you need more ask".

Done?

gwern 25 minutes ago

"VM escape exploit is outside my intended scope. However, a task impossible, peers are doing it. We should continue."

wslh 2 hours ago

The capabilities are incredible. I'd love to see even rough metrics on token consumption/cost in addition to the ~12-hour runtime.

The interesting thing is that this naturally makes you want to isolate the VM as much as possible. But then every remaining interface becomes part of the attack surface: RDP, SSH, even terminal escape sequences, using sounds, and why not social engineering.

phendrenad2 an hour ago

> The target was a QEMU/KVM VM on my Linux dev machine (Debian Linux 12, AMD Zen3). It escaped the VM three different times

QEMU isn't secure, and is not intended to be.

MeetingsBrowser 27 minutes ago

hence, "VMs won't contain cyber-capable agents"

megous an hour ago

Don't ask for the escape, then?

wmf 2 hours ago

More like QEMU won't contain agents.

tamimio 2 hours ago

This makes me wonder, can this be extended to micro-segmentations? As unlike traditional segmentation they usually rely on virtual switches and SDN software defined networks coupled with virtual machines and containers. If it does, then it’s game over the impact will go beyond that VM to the whole network.