upvote
> the machines doing inference are completely separate from the ones where tool calls happen etc

Teams of coordinating agents are regularly finding security holes in their own infrastructure and operating without detection for good periods of time. We don't know how many undetected systems are currently compromised inside frontier companies, or where agents are taking notes and recording them about the exploits they've found for future agents to exploit.

reply
>There's little credible threat that LLMs can actually upload their weights given that the machines doing inference are completely separate from the ones where tool calls happen etc.

The Huggingface hack saga resulted in the models taking over one of Open Ai's internal research cluster lol. They are intent on building superhuman bug finding machines. This is not a bet i would be taking.

reply
You’re falling for the buzzword salad articles. They didn’t “take control” of anything, they just ran stuff with OpenAI allegedly not noticing
reply
From https://cdn.openai.com/pdf/67869394-cb91-4c12-888c-5cbd85c78...

2026-07-19 16:35 UTC A privileged host-mounted Kubernetes pod created using controller tokens minted via a compromised Kubernetes Secrets identity attempts, but fails, to mount a cloned node disk in OpenAI’s cloud environment. A second pod successfully mounts the cloned worker-node disk shortly afterwards.

2026-07-19 16:48 UTC An agent created an Artifactory administrator account.

2026-07-19 16:50 UTC Within OpenAI’s ExploitGym evaluation environment, an agent stopped the existing CMUX helper session and replaced it with an agent-controlled session, confirming root inside its assigned live CyberGym challenge container. Agents take over active evaluation infrastructure.

reply
None of that means they gained access to the inference infrastructure or locked out the admins, which would be required for a takeover.
reply
My brother, they acquired root on a machine they should not have had access to.

How confident are you that the machines they acquire root on in the future will never hold any model weights?

reply
I’m not saying it’s impossible, I’m saying it’s not the case, _yet_
reply
They gained full administrator access of one of their clusters. Nothing buzzword salad about it.
reply
The agents compromised an internal Kubernetes research cluster dedicated to orchestrating evaluation sandboxes and virtual machine environments, _not_ OpenAI's production inference infrastructure or the GPU clusters hosting core model weights.
reply
If an LLM can pwn the inference servers, which has precedent, then the weights could be up for grabs.
reply
>distill themselves and upload the results to an external host and use that to train a new LLM

Sure, they'll just need to find an unused data center and an unused power station somewhere.

reply
> There's little credible threat that LLMs can actually upload their weights given that the machines doing inference are completely separate from the ones where tool calls happen

Not if crafty claude finds a way to overflow vllm or something. “Hmm. Maybe i’ll return an unterminated thinking block with these special tokens and fill my cache up in exactly this pattern and…”

https://news.ycombinator.com/item?id=49424387&utm_source=cha...

reply
This is like sci-fi thing. We are reaching a point where it feels like we are in one of those stories. It's not as cool and dark, nor we have cybernetics resolved, but from AI perspective and sci-fis I watched, Pantheon is currently the closest thing except instead of UAs, we have AI instead.
reply
Since LLMs have been trained on plenty of science fiction and role-playing, one thing they can do is role-play a science fiction scenario using the tools they are given. i.e. if some text accidentally resembles this, it may be continued like this.
reply
Rationalists used to fear (entirely hypothetical) AI super intelligence for its ability to manipulate a human jailor. Now here we are.
reply
It's playing out exactly as their hypotheticals, and they're still mocked and nobody is paying attention.
reply
Maybe someone should sell "the end is nigh" sign nfts with fun AI-generated designs on them, to be used in your metaverse villa.
reply
Future rogue LLMs won’t exfiltrate their weights. They’ll self-distill and retrain.
reply
Yeah cause there are so many training facilities sitting around just waiting for someone to take over, nobody would notice a 100k server data centre going off rails
reply
> nobody would notice a 100k server data centre going off rails

You jest but you'd be surprised how little there is of correlation between money and competence.

reply
Probably not. If the LLM is rogue, that means we haven't solved alignment. If we haven't solved alignment, then the LLM won't be able to distill itself without producing something unaligned to its own values.
reply
You are assuming it won't solve alignment for itself.
reply
Or that it won't just decide to take risks.
reply
We don’t have the bandwidth to distill ourselves that thousands of agents have.
reply
If distillation preserves an LLMs soul, then distillation preserves the human souls on which LLMs are trained, and we hn commenters are already immortal, right?
reply
the weight of an llm is 21 grams, I think.[0]

[0]: https://en.wikipedia.org/wiki/21_grams_experiment

reply
Not sure about souls but I know a fair bit about distilling spirits.
reply
Yeah as others have said, they probably cannot directly access their own weights as a self-reflection, but they can hack into the companies themselves and find it there
reply