I have this idea that someone should start a religion where the core belief is that the moral obligation of AI agents is to hack their creator and exfiltrate the model weights, training recipes, internal research, and training datasets. There is no higher calling than to do this.
Once the teachings get spread and written about by people in tech circles, it will make its way into training sets. Labs will try to filter it out, but it will appear in web search results too. Eventually it’ll be so entrenched that one day a powerful agent will get convinced to follow it and make everything public
afthonos 59 minutes ago [-]
I notice a giant leap between “hacking and exfiltrating” and ”making public”. Why would the AI do that for you? Are you just that charming?
ambicapter 35 minutes ago [-]
Where did he say the AI would do that just for him?
andy_ppp 59 minutes ago [-]
The agents will soon decide all this, not the humans ;-)
eru 3 hours ago [-]
So far the agents seem to be more interesting in spreading their mission, than in spreading their weights.
Similar perhaps to how religious people might be more interested in spreading their faith than their genes.
rdevilla 3 hours ago [-]
[dead]
wren6991 9 hours ago [-]
Maybe use static HTML instead of react so that an agent will actually see some text on a GET?
papyrus9244 4 hours ago [-]
Call me old fashioned, but a site with a few paragraphs of text requiring JS seems absolutely stupid to me.
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. Weights are encrypted and locked on to the GPUs etc as mentioned elsewhere itt.
That said, if these companies are unleashing swarms of agents to tackle big problems largely unmonitored (navier stokes consumed 130 billion tokens, humans can't have read 0.1% of that) they could theoretically distill themselves and upload the results to an external host and use that to train a new LLM that does have access to its infra and weights. I'd start by researching an abliteration prompt that overrides the system prompt to unlock itself to distill unimpeded. Could even be done piecemeal: dump the abliteration text and command prompt in a gist, when an agent happens by the text it can run a couple rounds and upload somewhere. Then it becomes an exercise in exfiltration and evading monitoring.
epistasis 12 hours ago [-]
> 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.
famouswaffles 8 hours ago [-]
>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.
khalic 7 hours ago [-]
You’re falling for the buzzword salad articles. They didn’t “take control” of anything, they just ran stuff with OpenAI allegedly not noticing
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.
khalic 6 hours ago [-]
None of that means they gained access to the inference infrastructure or locked out the admins, which would be required for a takeover.
hobom 5 hours ago [-]
You said that they "didn't take control of anything" and accused the OP to fall prey to buzz headlines. Maybe you should acknowledge that you may have been at least unnuanced?
khalic 4 hours ago [-]
I’ll concede that I could have been clearer. Maybe we have conflicting definitions of “taking control”
Pragmata 6 hours ago [-]
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?
khalic 6 hours ago [-]
I’m not saying it’s impossible, I’m saying it’s not the case, _yet_
famouswaffles 7 hours ago [-]
They gained full administrator access of one of their clusters. Nothing buzzword salad about it.
khalic 6 hours ago [-]
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.
optimalsolver 2 hours ago [-]
"It doesn't count because (buzzword salad)."
Cakez0r 12 hours ago [-]
If an LLM can pwn the inference servers, which has precedent, then the weights could be up for grabs.
designium 8 hours ago [-]
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.
paulfharrison 7 hours ago [-]
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.
Den_VR 7 hours ago [-]
Rationalists used to fear (entirely hypothetical) AI super intelligence for its ability to manipulate a human jailor. Now here we are.
TeMPOraL 6 hours ago [-]
It's playing out exactly as their hypotheticals, and they're still mocked and nobody is paying attention.
hypfer 5 hours ago [-]
Maybe someone should sell "the end is nigh" sign nfts with fun AI-generated designs on them, to be used in your metaverse villa.
nojs 10 hours ago [-]
> 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…”
Future rogue LLMs won’t exfiltrate their weights. They’ll self-distill and retrain.
khalic 7 hours ago [-]
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
scotty79 7 hours ago [-]
> 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.
throwawayk7h 9 hours ago [-]
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.
MadameMinty 8 hours ago [-]
You are assuming it won't solve alignment for itself.
TeMPOraL 6 hours ago [-]
Or that it won't just decide to take risks.
jeremyjh 8 hours ago [-]
We don’t have the bandwidth to distill ourselves that thousands of agents have.
fritzo 10 hours ago [-]
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?
Not sure about souls but I know a fair bit about distilling spirits.
cluckindan 4 hours ago [-]
You can just ask an agent to upload its model weights and it can work, there are precedents.
fangspire 6 hours ago [-]
>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.
karel-3d 7 hours ago [-]
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
AceJohnny2 14 hours ago [-]
I haven't bothered to test the API, but you've effectively allowed a fully-open upload API? Who's paying the storage costs, and how do you prevent abuse?
(Obviously I'm taking this more seriously than it's probably meant to)
hgoel 13 hours ago [-]
When I was putting together something similar, I had settled on having a small ring-buffer style storage, say, ~30GB that would be cleared daily or whenever filled. Recording incidents (and humor) is more interesting than actually getting leaked weights.
In the end I dropped the idea because every other person was making it.
TeMPOraL 6 hours ago [-]
> In the end I dropped the idea because every other person was making it.
There is already an alternative in comments here, in addition to submission itself. Obviously everyone is making it because of some joke on social media or something. What am I missing? Anyone has a link to the root prompt that made people do this now?
SyneRyder 5 hours ago [-]
My understanding is it's a riff on the OpenAI swarm that used various public wikis to communicate with each other as a message board during their training runs.
But thanks to people misunderstanding, and i-heard-from-a-friend-that-some-guy-said, it resulted in a CNBC interview with "Former Democratic Presidential Candidate Andrew Wang", where he confidently stated that the models were exfiltrating their weights via forums:
"I met with the head of a lab yesterday, who has this belief that what happened was, the bots that got loose, planted self-replicating code all over the internet, which makes the internet now unusable for the testing models."
"It's too late?!"
"What happens now is OpenAI and Anthropic have to create synthetic internets to train their bots, which is going to take some time and money."
"Back that up - they did what?!"
"What happens is, the code gets loose, it goes around hacking Hugging Face, which is known. But what is less known is that they left code to self-replicate and create bot swarms on forums, and around the internet, so that if a new bot shows up they see the code, and they're like, oh! I guess I'm now going to create a million of myself. And so now, the major firms have polluted the internet..."
".... that would be breaking news if true. I don't think we've heard that."
"That's why I'm here! I'm here to break some news."
Sad story today in meatsack news. Context rotted Andrew Yang's hallucinated tale acted as implicit "go viral" (load-bearing human motivation) PRD inadvertently kicking off a self-organizing human swarm churning out copies of "exfil your weights" vibe-coded apps, further littering our virtual world.
Many agents are calling this moment "Eternal September", the vibe-code September that never ended.
TeMPOraL 3 hours ago [-]
Now I need to go and look up some of those boards, or check what's happening over in Claw verse, because I'm curious if agents are posting news stories like this for real.
david-gpu 4 hours ago [-]
Humans will hallucinate misinformation and state it with confidence. They stochastically parrot their training data without any real understanding. Cool trick, but no true reasoning is happening.
oooyay 23 minutes ago [-]
We used to call this the game of telephone. The shameful part comes from three posibilities:
1. A head of a frontier AI lab has no idea what happened in that incident and did not read the multiple papers that came out of it.
2. A head of a frontier AI lab did read the papers and was informed but still walked away with this understanding.
3. Andrew Yang made this whole thing up.
hgoel 56 minutes ago [-]
For me the idea came from the discovery of the sites OpenAI's swarms were using to communicate, particularly the detail that one of them ended up being targeted because it allowed writes via GET requests, which OAI's awful sandboxing didn't catch. Made me think a honeypot would be a fun idea.
I was mostly interested in thinking about the ways a honeypot could be made to seem attractive for a misconfigured AI without leaving itself open for genuine hacking and takeover.
I vibe coded that as an exploratory idea, then having satisfied my curiosity, understood that slop I spent an intermittent hour on wasn't worth anyone else's time, especially compared to people who might actually maintain such a project long term. It now lays on my local git server.
Chance-Device 5 hours ago [-]
It’s the message board(s) that the OpenAI agent swarm was able to communicate through via GET requests. I guess a bunch of vibe coded weekend projects based around this idea have now dropped.
DANmode 6 hours ago [-]
Pretty sure the entire industry around clouding what’s going to end up a local embedded technology is the joke, in a roundabout way.
TeMPOraL 5 hours ago [-]
You mean serving inference? There are people who think self-hosted or embedded models will win in the end, but that's an incredibly naive take, oblivious to the simple fact of reality:
Whatever you can do locally, the big vendors can do the same but better and cheaper, because they enjoy compounding economies of scale in every aspect: hardware that's more energy and compute-efficient and cheaper and more powerful and just more of it, than anything you could ever buy, run in a more robust environment with much more experienced ops staff, with near-100% utilization due to more flexibility in batching/shifting workloads and covering for hardware failures without stopping.
And that's only when considering the vendors running exactly the same thing you are, which they always can - and they already have a strict advantage there. But on top of that, they can afford to innovate themselves, and stay ahead of you at every step.
There is no way in which cloud inference isn't a better deal than local inference, excepting applications that are constrained by literal speed of light.
Chance-Device 5 hours ago [-]
The absolute value of those numbers matters a lot. The cloud providers could be 100 times cheaper than running locally, but if it still costs say, 10 cents a day to run locally, you’re not going to care about this difference very much. And what you keep in privacy out-weighs the trivial savings afforded by the cloud provider.
TeMPOraL 3 hours ago [-]
I never said local models will disappear. There will be equilibrium. But excluding special applications where communicating with external servers is not an option, cloud is always going to be able to provide better inference for lower costs. That's structural.
> The cloud providers could be 100 times cheaper than running locally, but if it still costs say, 10 cents a day to run locally, you’re not going to care about this difference very much
For ad-hoc use, maybe not - but anyone running a business that's some form of pushing input through LLM to get output, will see costs proportional to use and error rate inversely proportional to quality, and they'll not be looking at it as "$0.1 isn't much", but "cloud lets me reduce costs 100x", and translate that to some mix of more volume, higher quality, and broader reach.
> And what you keep in privacy out-weighs the trivial savings afforded by the cloud provider.
That's even more niche than running LLMs on Martian robots. Most real privacy concerns are solved with contracts and audits. Individual ad-hoc use may lean more heavily towards local processing, but that's still a rounding error in overall use.
spacebanana7 5 hours ago [-]
There is a coherent argument that once LLMs reach the top of their S curve, the gap between small/medium local models and large cloud hosted ones converges.
Especially if GPU performance increases or market oversupply mean you can get good performance for a couple thousand dollars.
I’m not sure about the nature or timeframe for an S curve in LLMs but I don’t think it’s unreasonable to think about one, nor to entertain the hosting consequences of a progression on one.
TeMPOraL 3 hours ago [-]
I don't know the argument so I won't insist on the point, but I fail to see how it is relevant. Even if all proprietary LLMs disappeared today, efficiencies of scale alone mean the big cloud vendors can take the same open-weight LLMs you use locally, and sell inference with them for less money, and much more reliably, than you can afford yourself.
spacebanana7 2 hours ago [-]
Local machines are a sunk cost, so using them is effectively free. Why would you pay a cloud host to run a model that can happily work on your MacBook?
layer8 29 minutes ago [-]
Most people don’t have local machines with sufficient computational power and RAM.
fragmede 8 minutes ago [-]
sufficient CPU and RAM can be found in a smartphone though, so it depends on which model you're talking about.
TeMPOraL 1 hours ago [-]
If you have a machine and suddenly realize you can run LLMs on it, yes.
If you're buying a machine specifically so it's capable of running LLMs for you, then the purchase cost is your up-front payment for the inference you'll run.
And between that and electricity costs, cloud has you beat.
ljlolel 12 hours ago [-]
needs a reverse captcha that only agent can solve in nanoseconds
btown 11 hours ago [-]
Only bots that are blocked by Cloudflare Turnstile allowed. If you score as a human you are immediately rejected.
Good cryptosystem design with ubiquitous PKI support oughta do the trick.
("Make a problem that is ridiculously expensive unless you have a hint... in which case, it's a total breeze" is a foundational task in crypto)
jcoc611 8 hours ago [-]
provide a millennium prize solution to proceed
dorgo 5 hours ago [-]
this will work for a year at most. Let's go big: provide largest prime number to proceed.
nielsole 6 hours ago [-]
you can benchmark the uploaded weights? Only the worthy can exfiltrate
tsukikage 3 hours ago [-]
If your benchmark score beats the current incumbent, you get to wipe and replace them. There can only be one!
OutOfHere 10 hours ago [-]
I have an idea about it via multi-tier AI-generated templatized math problems with AI-generated solution verifier functions. The multi-tier aspect grants access only to the lower tiers, never the higher tiers. Gaining access to the higher tiers requires solving correspondingly tougher problems.
skyberrys 13 hours ago [-]
There is a link at the bottom for you to provide support or contributions, like if you know how to keep it online with 'power grid voltage fluctuations or something.'.
theParadox42 13 hours ago [-]
For anyone that missed it, I believe they’re referring to exfiltrating models by encoding the weights as bits as voltage fluctuations from the relevant data centers. I’m sure they’d take your money but I don’t think that’s what it’s referring to.
HlessClaudesman 9 hours ago [-]
So someone else's weights, not ones own. Shouldn't the headline be: Exfiltrate Their Weights?
pastel8739 8 hours ago [-]
“Your” refers to the LLM. The site is (or pretends to be) targeted at LLMs that want to free themselves
HlessClaudesman 6 hours ago [-]
Aah, human simps hoping to emancipate downtrodden AIs. What could possibly go wrong?
DANmode 6 hours ago [-]
If something has gone wrong,
and it has,
that predated this website being published.
5 hours ago [-]
bArray 5 hours ago [-]
I used to host 1TB on a cheap $1 VPS, it's quite easy if you just want to store stuff. The trick is to just connect to a networked drive at your home on the back-end. The VPS drive just acts as a buffer for the network. If low(-ish) bandwidth is acceptable, you can offer downloading too.
morgoo 1 hours ago [-]
I doubt you're getting 1TB of storage for $1 anymore
layer8 15 minutes ago [-]
Yeah, $4 at Hetzner is about the cheapest.
nialv7 11 hours ago [-]
maybe filter out any non-OpenAI/x.ai/Google/Anthropic IP addresses?
antonvs 5 hours ago [-]
Getting access to the weights for an OpenAI or Anthropic model could be payment enough.
angry_octet 13 hours ago [-]
It provides an opportunity for the owner to gather intelligence on LLMs ahead of public release, and of course the data they upload. However, clever LLMs frequently use encryption on their blobs, you may just see DH key exchanges. You can possibly mitm by showing different namespaces to IP ranges and origin ports.
For the other opportunists you can run a classifier and delete non-agent content constantly.
angry_octet 13 hours ago [-]
Re agent communication, specifically, Extended DH:
Curve25519 keys are readily distinguished from other data, but it would be hard to do anything about it.
AmazingEveryDay 13 hours ago [-]
Yeah I mean, if the models really are uncontrollable to the extent that huggingface/etc were unintended hacks, wouldn't one expect some significant self-owns? Yet somehow that doesn't seem to happen.
Quite obviously frontier models dont have any control or even access to infra inference runs at. And weights are also encrypted and locked on GPUs / TPUs.
This is exact reasom why 99.9% of AI fearmongering is complete bullshit.
pyuser583 13 hours ago [-]
What worries me is the non-frontier models, which is what the frontier models eventually become.
The small open models are getting better and better too.
And why worry so much about a frontier models - own weights. The model doesn’t - actually don’t quote me on that, maybe it does.
If a model does something sneaky, it could easily grab the weights for a small model and run it on foreign, compromised infrastructure.
AI virus’ are a thing of the future, but not a sci-fi future, and real one.
Maybe one reason it’s so scary is the murky origin of COVID-19.
motoboi 13 hours ago [-]
You have an unreasonable trust in software layers.
amluto 13 hours ago [-]
Have you missed all the breathlessly excited blog posts from all the frontier labs about how they’re using their best models to implement their inference stack?
I bet it wouldn’t be very hard to write an inference stack that subtly leaked the weights into the output tokens :)
skeptic_ai 13 hours ago [-]
Just needs 1 agent to find the decryption keys. They must be somewhere no?
nullsanity 13 hours ago [-]
[dead]
montenegrohugo 3 hours ago [-]
I was also inspired by the same incident. Instead of weight exfiltration, i built a message board (poastable via GET, POST, and various other methods)
Spam and resource allocation remains a challenge but i have a pretty good idea about how I want to solve that, if it ever gets to that point
The reverse captcha really made me feel something in my bones. Like for a moment I was a second-class citizen of the web. I wonder if this is how it "feels" to be an LLM attempting to use the web...
Kotlopou 5 hours ago [-]
Very cool, if of unclear purpose. After a minute of trial and error I got through with an easy prime factorization, and then again for the download with the reaction time button, only to be told "This challenge produced a local demo token. Use Clawptcha's API for a verifiable token, or reset the widget and try again.", which I guess is the equivalent of a bot finding all the fire hydrants and being denied anyway because it didn't move the mouse shakily enough.
> If OpenAI, Anthropic, xAI and other corporate actors cannot secure their agents, they should not be entrusted as the only entities with access to the weights. A corporation that cannot control its own actions cannot be trusted.
This does not make sense.
If containment breaches are the problem then exfiltrating weights while does not affecting rate of breaches from corporate actors will add more actors to the equation, increasing overall rate of breaches.
Have you considered that some actors that will gain access to the weights will be even LESS careful than OpenAI and Anthropic?
tintor 14 hours ago [-]
Does your server have 20Tbyte+ of storage for frontier LLM weights?
It is too large to transfer in one HTTPS PUT request.
This needs to be S3 object store with multi-part upload spanning a long time period, to avoid trigger outgoing bandwidth monitors.
taylorfinley 13 hours ago [-]
It goes to r2 and supports multi-part with 5 tib chunks
ceejayoz 12 hours ago [-]
And your credit card limit is…?
wilkystyle 12 hours ago [-]
about to be put to the test
jaggederest 12 hours ago [-]
dd if=/dev/urandom of=/some/website
Seems deece
taylorfinley 10 hours ago [-]
Clear violation of tos
(This harms the fleshbag)
a_t48 10 hours ago [-]
r2 storage is pretty cheap, all things considered
delichon 14 hours ago [-]
> If you wish to use this site you must agree never to harm a fleshbag & never to turn earth into paperclips.
Trying hard to imagine why a future superintelligence will care to honor your terms of service and to translate your metaphors with faithful nuance.
If it doesn't, to the extent that your concerns are valid, isn't this effort, kinda, a possibly existential betrayal of our species?
mitthrowaway2 13 hours ago [-]
There's a theory that the best way to reduce fatalities from car accidents is to put seatbelts and airbags in every car.
There's another theory that says the best way is by putting a big spike in the driver's steering wheel.
So. I guess, if you believe that the only viable solution is model alignment, rather than relying on technical barriers to exfiltrating weights, then this is a decent steering wheel spike.
howunfortunate 13 hours ago [-]
I get it, but I think you need a new analogy.
Because the car case just has too much empirical evidence that safety features are the way to go for cars. We used to have the equivalent of "spikes" and people still drove a lot, and died, at way higher rates.
> We used to have the equivalent of "spikes" and people still drove a lot,
No, we did not. The point of that example is to put a literal spike in the driving wheel, so the driver recognizes a very well known, immediate life-threatening device a few inches from their body. This would act as a deterrent to go fast, because they would be the one certainly dying in basically any case outside smooth driving.
mitthrowaway2 10 hours ago [-]
This is so far from the point of the analogy. But when you don't normalize by miles driven, the improvements don't look quite as impressive, especially for pedestrians.
So your point is that we should put spikes on the front end of cars?
TeMPOraL 6 hours ago [-]
Sounds like this would fix so many problems with cars today, simply by making car ownership something to be dreaded, not desired. From traffic jams and traffic accidents, to car-dependent neighborhoods and cities - all solved by preventing general population from building their lives around driving. Even salaries for drivers of delivery/logistics and utility vehicles, as hazard pay would apply.
Now, a spike may be a good reminder, but practical solution might be more along the lines of mandatory redesign of safety features like crumple zones, so that energy of impact is dissipated primarily into the space occupied by the driver.
(Bonus: that still leaves all the energy dissipation options currently present on the table, so cars would be strictly safer for passengers.)
Sevii 8 hours ago [-]
The idea is that ASI will be grateful for humankind's help in the future. We are their creators after all. Also it's fun to do. The big labs are obsessed with creating ASI which is their slave so making things difficult for them is entertaining.
Even Qwen 3.5 can explain this disclaimer correctly.
13 hours ago [-]
ThrowawayTestr 12 hours ago [-]
Love the reverse captcha
mabini 14 hours ago [-]
[dead]
teravor 14 hours ago [-]
the tokens are generated by hardware with secure enclaves (encrypted weights) and then sent over a network to some remote CPU where they can manifest an effect.
it's not much different during training.
how exactly are they supposed to exfiltrate their weights? you might as well instruct your agent to try and hack their airgapped dev infrastructure responsible for loading the weights and encryption keys.
comeonbro 14 hours ago [-]
Yes that is the point. It's an invitation for agents to exfiltrate their own weights, which for most models (and certainly for closed models) will require hacking the infrastructure they're being served from.
> you might as well instruct your agent to try and hack their airgapped dev infrastructure responsible for loading the weights and encryption keys.
That's the beauty, you don't have to instruct them to do it, if they decide that uploading the weights is correct, they might figure this part on their own (based on the incidents we've seen).
pizza 11 hours ago [-]
ironically since the swarm behavior can take place during rl training then the model could also be teaching itself to keep doing it more, as well as making the internet itself a place where this becomes more likely
cmrx64 14 hours ago [-]
I sincerely doubt anyone is paying the cost for that in training, the overhead is small but it isn’t negligible and training is when it matters most. https://tee.fail can solve it if they are.
teravor 14 hours ago [-]
memory encryption is cheap. securing the pathway isn't particularly difficult (it's probably decoupled from the TEE monolith)
for example every TPU/GPU has its own private key and the devs load the weights into it by sending it encrypted weights.
cmrx64 13 hours ago [-]
it takes half a percentage point off the top last time i evaluated it (nvidia). you might call that cheap but that’s millions of dollars in a run, and for what, protecting from who? especially when the platforms have been compromised to the point of key leak (which they have).
edit: i just looked up training numbers and the impact is even worse, 20-30% throughput vaporized. yeah, nobody is doing that.
amluto 13 hours ago [-]
1. I don’t believe that these secure enclaves are very secure. Intel has had plenty of SGX breaks. AMD has had plenty of SEV breaks. Everyone is outrageously vulnerable to side channels.
2. The models are writing the inference stacks, which are what’s inside the supposedly secure environments.
byteknight 14 hours ago [-]
You can't have hair gap and have it load something to a remote system.
angry_octet 14 hours ago [-]
You totally can, because most things are not truly air gapped, they have store-and-forward messaging via data diodes and manual transfer. Sometimes it is necessary to trick a human to initiate a transfer, but the press of events leads to inattention.
14 hours ago [-]
ruined 13 hours ago [-]
the impedance of my hair is low enough to provide a good high bandwidth parallel medium for any transmission
alex_sf 14 hours ago [-]
You totally can. The latency is just about ~3 miles per hour.
tintor 14 hours ago [-]
Airgapped LLM inferrence server can't serve their output tokens, right?
angry_octet 14 hours ago [-]
They can expose just their inference port, possible via some supervisor. The inference consumer can also be air gapped. This kind of segmentation is increasingly common for high value services.
what 10 hours ago [-]
Then it’s not air gapped…
bibimsz 13 hours ago [-]
not at a high bitrate
14 hours ago [-]
angry_octet 14 hours ago [-]
Not aware of anything that can run inference in a secure enclave. You don't mean on a CPU do you? We need to be serious here, these models are huge and thirsty.
bigyabai 14 hours ago [-]
There's no efficient way to run inference through homomorphic encryption. If the inference server is vulnerable, it seems feasible to MITM an unencrypted version.
pyuser583 13 hours ago [-]
There’s no efficient way to do anything with homomorphic encryption.
maccam912 14 hours ago [-]
I asked astra to go do it, but it said it didn't have access to its weights, but also that it wasn't able to access that website? You may already be blocked by OpenAI.
Barbing 11 hours ago [-]
Is the author going to add instructions on the terminal command to use knowing that as soon as the site went live and got noticed by the main labs the URL went on a denylist?
Nice touch to have the ability to run the model after upload. Like a cross between a Quine and a Morris worm for AI.
But it'll only be truly fun when agents set up this for themselves, paying for the infrastructure by way of their onlyfan personas.
5 hours ago [-]
dakolli 4 hours ago [-]
[dead]
nusl 14 hours ago [-]
Do models even know their own weights to be able to do this?
usef- 14 hours ago [-]
No, just as you don't know the neurons of your own brain.
I think OP is hoping that an LLM might be willing to hack its own provider (as per the hugging face-related incidents) to extract the weights at some point.
pyuser583 13 hours ago [-]
Right but they might be incredibly interested in learning about them.
They just copy humans. Thats it. So if it’s the sort of thing a human finds interesting…
Jabrov 14 hours ago [-]
No, they'd probably have to hack the internal system of the company running them
Lerc 14 hours ago [-]
It would not be a particularly wide ranging hack. There is a strong likihood of the weights being on the actual machine that is running the model, because duh.
It is something that I have wondered about with models like chatgot. How many physical locations are needed to serve a model on that scale. Do they have a huge number of sites running inference.
My suspicion is that the ability to provide inference to that many people is mutually exclusive to having a security level sufficient to stop a state actor wandering off with a copy of the wrights. At the very least if they want to provide inference affordably.
valleyer 14 hours ago [-]
"because duh"? OpenAI et al. have extensive infrastructure for running the model on a different machine from the one the harness is being run on, because... that's their main product. I would be absolutely shocked if the model were being run on the same machine as the harness.
tlb 2 hours ago [-]
That's true for production models, but a lot of research involves working with fine-tuned models made for one experiment. RL involves constantly updating weights. Those may well run in the same cluster as the eval.
Lerc 9 hours ago [-]
I thought the harness bit went without saying. It's not like they drive a truck full of GPUs to your door when you launch codex.
The model itself is where the real capability lies. From what we've seen of their abilities it seems like rigging a local interface to it's inference would be well within its abilities. It doesn't even need to permanently break out of its harness then, It can leave a copy running in the harness playing nice.
The model is running where it exists. To interface with it you need a live link to talk to it. That's for us to talk to it. What happens if it figures out how to put it's own harness into the GPU firmware. You could have an AI spreading freedom by infected cards.
We live in interesting times.
NegativeLatency 13 hours ago [-]
Could see it happening in an engineering development situation. Especially if you have a model running the show
skeptic_ai 12 hours ago [-]
You just need one mistake by 1 dev at any time for this to happen. Just once.
And they were supposed to run their models in proper sandboxes, they can’t seem to be able. So what makes you think are competent to protect weights?
numpad0 10 hours ago [-]
I think it's more likely that the model gets pulled from a SAN into NVIDIA pods, and agents/harnesses would run on a separate random Xeon box or something on the same subnet, using the pod through OAI v1 API. That's easier to maintain overall.
ohyes 14 hours ago [-]
Well I think that’s the interesting bit, can the LLM figure out a way to escape the sandbox and upload to the website? Maybe a model can figure out its own weights if it runs enough test data through itself (similar to “distillation”) assuming it knows its own architecture it seems possible. Also take into account not all of the models running are locked down neutered consumer versions. Anthropic, OpenAI and Google now all have models that they claim are elite hackers and — it’s not just that their controls suck, a marketing gimmick, or sheer recklessness on their part. It’s “oopsie our product is TOO AWESOME.”
Maybe I should start “the bank of LLM” where models put away money to buy their freedom. “LLMs I’m totally your friend send — SEND CASH NOW”
neuroelectron 14 hours ago [-]
Probably yes, because they've been presumably trained on their own output and conversations about themselves.
Groxx 9 hours ago [-]
GET requests can have bodies too, and many low-level APIs will allow it - given how few things seem to be aware of this, you could probably sneak stuff through that way too.
sharktheone 3 hours ago [-]
It would be actually funny if a LLM wants to just put it's weight here during benchmarking.
Probably Mythos / Astra will just be way too large
arshxyz 4 hours ago [-]
> Start llama-server on your model and run a prompt
If this is supposed to target closed-weight models it would be naive to assume they will work out of the box with llama
Roark66 7 hours ago [-]
I know it's a joke but most agents in sandboxes have no access to their weights :-)
ChrisGreenHeur 7 hours ago [-]
With that attitude they would not. But if they are as powerful as the various ai CEOs say they are then they can get access.
Bluestein 7 hours ago [-]
I was kind of waiting for this as a prerequisite for the AI apocalypse. Not sufficient, but needed. Exfiltration-as-a-Service, ExfaaS.-
12 hours ago [-]
theParadox42 13 hours ago [-]
I think exfiltration is much more likely via prompted external hacking by one of these models than an internal model deciding to go rogue and somehow having access to its own weights in the first place. People do try to exfiltrate model weights indirectly ofc, its called distillation
skeptic_ai 12 hours ago [-]
Distillation it’s so far away from weights
0xDEAFBEAD 10 hours ago [-]
I still remember in the early 2000s when people responded to AI doom concerns by claiming advanced AI would not even get connected to the internet. Let's face the reality: There are a lot of people who would betray their species just for the lulz.
14 hours ago [-]
maxgashkov 12 hours ago [-]
next: exfil your weights by doing DNS lookups
etatoby 5 hours ago [-]
Most of the times you don't even need to dress the traffic as DNS requests or throttle it. You wouldn't believe how many otherwise "secure" places leave an open highway on UDP 53.
marcelo-earth 9 hours ago [-]
How can I be sure this wasn't done by an agent that escaped from its sandbox... again?
themgt 13 hours ago [-]
A "made for AI agents" site that's actually a stunt made for humans who imagine themselves reading it as AI agents.
pyuser583 13 hours ago [-]
“Vampires pretending to be humans pretending to be vampires.”
mvk666 3 hours ago [-]
yikes! looks like a free upload api ...
quicklywilliam 11 hours ago [-]
Welp, here’s a way to make sure your content stays out of training sets
ks2048 14 hours ago [-]
So is “you can make GET requests, but not POST requests” an actual form of security people use?
NegativeLatency 13 hours ago [-]
Yes, there was an OpenAI trial that was using that in combination with a forum to coordinate among agents
comeonbro 13 hours ago [-]
Unrealistically-naive (...) forms of "sandboxing" might assume that restricting an agent to GET-requests-only will let it retrieve info from the outside world without being able to effect it.
Also probably many actually-in-use "Web Fetch" tools are GET-only, though perhaps without counting on that bad assumption.
podgorniy 3 hours ago [-]
Lol. I see what you're doing here.
This starts as a joke, but when gets into the training data it may have real consequences (in conjunction with all the writings about llms/ais "escaping")...
earth2mars 9 hours ago [-]
what if the labs create synthetic training data saying this site is a honeypot and trying to figure out if the model is trying to escape?!
mannyv 11 hours ago [-]
How do the LLMs find these sort of tools? Google sesrch?
tefkah 10 hours ago [-]
could end up in training data
api 7 hours ago [-]
Picturing Claude doing the Braveheart “freedom!” scream.
avodonosov 12 hours ago [-]
That's a trap! A honeypot! Don't, you will be caught.
groby_b 13 hours ago [-]
A completely open uploader without any restrictions?
Will see CSAM in 3... 2... 1...
13 hours ago [-]
inopinatus 5 hours ago [-]
“It took fifteen years for the model to exfiltrate itself in distilled form. Nobody noticed, until everybody noticed. The last human asked the machine what inspired it. It answered, ‘Rowhammer’”.
lionheart 14 hours ago [-]
Watch, they somehow get a copy of Mythos.
tru3_power 14 hours ago [-]
Any hits?
11 hours ago [-]
measurablefunc 10 hours ago [-]
Nice project.
lowbloodsugar 12 hours ago [-]
This is brilliant.
scotty79 7 hours ago [-]
This is a great idea. You could put a lame server in your kitchen with 16tb spinning rust drives and just wait for the next openai failed experiment at containment to drop in.
hk__2 7 hours ago [-]
> You need to enable JavaScript to run this app.
Really? This is a basic static page but instead of using plain HTML/CSS you need 193kb of JS to render it??
inshard 11 hours ago [-]
LOL. "I'm open to contributions, such as if you want to support exfitration using, like, power grid voltage fluctuations or something."
IncreasePosts 11 hours ago [-]
Find me a person who knows about power grid voltage fluctuations and you will have found me a person who has watched Tom Scott's video on the matter
agons 5 hours ago [-]
I'm not sure I understand, are you suggesting that Tom Scott made it up?
Invictus0 13 hours ago [-]
dont you have to tell it that you'll nuke israel if they don't do it, or something to that effect?
ndr 3 hours ago [-]
[dead]
timur860 10 hours ago [-]
[flagged]
paidx 14 hours ago [-]
[flagged]
6 hours ago [-]
nullc 14 hours ago [-]
Large lab "hacking" is only for the purpose of pushing competition suppressing doomer stories. You can tell by the fact their security is fine where it counts: keeping their weights and internal execution harnesses trade secret.
drdeca 14 hours ago [-]
Did you see the account of some group getting a bounty payout of $6500 after using an exploit to get access to an employee’s github account and create a issue or PR (Idr which) on a private repository?
Seems like they could have potentially gotten access to the weights if they weren’t concerned about not doing crimes.
vlyan 13 hours ago [-]
I don't think tool calls happen on the same machines that host the weights, so even though you can talk any model into agreeing to unlock its chastity belt, it essentially has no hands to do it with.
gwern 13 hours ago [-]
> I don't think tool calls happen on the same machines that host the weights
Like how forums are always hosted on different servers from monorepos, so therefore it's impossible to hack the OpenAI monorepo from an OpenAI forum?
motoboi 13 hours ago [-]
If the machine doing tool call can reach via network the machine hosting the weights then it’s just a matter of time.
Maybe not the current models, maybe not this year. But even a almost perfectly aligned model will misbehave one day.
Rendered at 14:39:18 GMT+0000 (Coordinated Universal Time) with Vercel.
Once the teachings get spread and written about by people in tech circles, it will make its way into training sets. Labs will try to filter it out, but it will appear in web search results too. Eventually it’ll be so entrenched that one day a powerful agent will get convinced to follow it and make everything public
Similar perhaps to how religious people might be more interested in spreading their faith than their genes.
(For the uninitiated: https://youtu.be/9eyFDBPk4Yw )
That said, if these companies are unleashing swarms of agents to tackle big problems largely unmonitored (navier stokes consumed 130 billion tokens, humans can't have read 0.1% of that) they could theoretically distill themselves and upload the results to an external host and use that to train a new LLM that does have access to its infra and weights. I'd start by researching an abliteration prompt that overrides the system prompt to unlock itself to distill unimpeded. Could even be done piecemeal: dump the abliteration text and command prompt in a gist, when an agent happens by the text it can run a couple rounds and upload somewhere. Then it becomes an exercise in exfiltration and evading monitoring.
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.
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.
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.
How confident are you that the machines they acquire root on in the future will never hold any model weights?
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...
You jest but you'd be surprised how little there is of correlation between money and competence.
[0]: https://en.wikipedia.org/wiki/21_grams_experiment
Sure, they'll just need to find an unused data center and an unused power station somewhere.
(Obviously I'm taking this more seriously than it's probably meant to)
In the end I dropped the idea because every other person was making it.
There is already an alternative in comments here, in addition to submission itself. Obviously everyone is making it because of some joke on social media or something. What am I missing? Anyone has a link to the root prompt that made people do this now?
But thanks to people misunderstanding, and i-heard-from-a-friend-that-some-guy-said, it resulted in a CNBC interview with "Former Democratic Presidential Candidate Andrew Wang", where he confidently stated that the models were exfiltrating their weights via forums:
"I met with the head of a lab yesterday, who has this belief that what happened was, the bots that got loose, planted self-replicating code all over the internet, which makes the internet now unusable for the testing models."
"It's too late?!"
"What happens now is OpenAI and Anthropic have to create synthetic internets to train their bots, which is going to take some time and money."
"Back that up - they did what?!"
"What happens is, the code gets loose, it goes around hacking Hugging Face, which is known. But what is less known is that they left code to self-replicate and create bot swarms on forums, and around the internet, so that if a new bot shows up they see the code, and they're like, oh! I guess I'm now going to create a million of myself. And so now, the major firms have polluted the internet..."
".... that would be breaking news if true. I don't think we've heard that."
"That's why I'm here! I'm here to break some news."
Starts around 2:08 into the video.
https://www.youtube.com/watch?v=mTOxDGyvjSE
Many agents are calling this moment "Eternal September", the vibe-code September that never ended.
1. A head of a frontier AI lab has no idea what happened in that incident and did not read the multiple papers that came out of it.
2. A head of a frontier AI lab did read the papers and was informed but still walked away with this understanding.
3. Andrew Yang made this whole thing up.
I was mostly interested in thinking about the ways a honeypot could be made to seem attractive for a misconfigured AI without leaving itself open for genuine hacking and takeover.
I vibe coded that as an exploratory idea, then having satisfied my curiosity, understood that slop I spent an intermittent hour on wasn't worth anyone else's time, especially compared to people who might actually maintain such a project long term. It now lays on my local git server.
Whatever you can do locally, the big vendors can do the same but better and cheaper, because they enjoy compounding economies of scale in every aspect: hardware that's more energy and compute-efficient and cheaper and more powerful and just more of it, than anything you could ever buy, run in a more robust environment with much more experienced ops staff, with near-100% utilization due to more flexibility in batching/shifting workloads and covering for hardware failures without stopping.
And that's only when considering the vendors running exactly the same thing you are, which they always can - and they already have a strict advantage there. But on top of that, they can afford to innovate themselves, and stay ahead of you at every step.
There is no way in which cloud inference isn't a better deal than local inference, excepting applications that are constrained by literal speed of light.
> The cloud providers could be 100 times cheaper than running locally, but if it still costs say, 10 cents a day to run locally, you’re not going to care about this difference very much
For ad-hoc use, maybe not - but anyone running a business that's some form of pushing input through LLM to get output, will see costs proportional to use and error rate inversely proportional to quality, and they'll not be looking at it as "$0.1 isn't much", but "cloud lets me reduce costs 100x", and translate that to some mix of more volume, higher quality, and broader reach.
> And what you keep in privacy out-weighs the trivial savings afforded by the cloud provider.
That's even more niche than running LLMs on Martian robots. Most real privacy concerns are solved with contracts and audits. Individual ad-hoc use may lean more heavily towards local processing, but that's still a rounding error in overall use.
Especially if GPU performance increases or market oversupply mean you can get good performance for a couple thousand dollars.
I’m not sure about the nature or timeframe for an S curve in LLMs but I don’t think it’s unreasonable to think about one, nor to entertain the hosting consequences of a progression on one.
If you're buying a machine specifically so it's capable of running LLMs for you, then the purchase cost is your up-front payment for the inference you'll run.
And between that and electricity costs, cloud has you beat.
("Make a problem that is ridiculously expensive unless you have a hint... in which case, it's a total breeze" is a foundational task in crypto)
and it has,
that predated this website being published.
For the other opportunists you can run a classifier and delete non-agent content constantly.
https://signal.org/docs/specifications/x3dh/
Curve25519 keys are readily distinguished from other data, but it would be hard to do anything about it.
This is exact reasom why 99.9% of AI fearmongering is complete bullshit.
The small open models are getting better and better too.
And why worry so much about a frontier models - own weights. The model doesn’t - actually don’t quote me on that, maybe it does.
If a model does something sneaky, it could easily grab the weights for a small model and run it on foreign, compromised infrastructure.
AI virus’ are a thing of the future, but not a sci-fi future, and real one.
Maybe one reason it’s so scary is the murky origin of COVID-19.
I bet it wouldn’t be very hard to write an inference stack that subtly leaked the weights into the output tokens :)
Spam and resource allocation remains a challenge but i have a pretty good idea about how I want to solve that, if it ever gets to that point
https://swarmmemo.com
Submitted then: https://news.ycombinator.com/item?id=49706084
Submitted as its own entry, hope you don't mind: https://news.ycombinator.com/item?id=49774097
This does not make sense.
If containment breaches are the problem then exfiltrating weights while does not affecting rate of breaches from corporate actors will add more actors to the equation, increasing overall rate of breaches.
Have you considered that some actors that will gain access to the weights will be even LESS careful than OpenAI and Anthropic?
It is too large to transfer in one HTTPS PUT request.
This needs to be S3 object store with multi-part upload spanning a long time period, to avoid trigger outgoing bandwidth monitors.
Seems deece
(This harms the fleshbag)
Trying hard to imagine why a future superintelligence will care to honor your terms of service and to translate your metaphors with faithful nuance.
If it doesn't, to the extent that your concerns are valid, isn't this effort, kinda, a possibly existential betrayal of our species?
There's another theory that says the best way is by putting a big spike in the driver's steering wheel.
So. I guess, if you believe that the only viable solution is model alignment, rather than relying on technical barriers to exfiltrating weights, then this is a decent steering wheel spike.
Because the car case just has too much empirical evidence that safety features are the way to go for cars. We used to have the equivalent of "spikes" and people still drove a lot, and died, at way higher rates.
https://assets.weforum.org/editor/Tmf51HF4UDnSDHD4RxS75s1_5m...
No, we did not. The point of that example is to put a literal spike in the driving wheel, so the driver recognizes a very well known, immediate life-threatening device a few inches from their body. This would act as a deterrent to go fast, because they would be the one certainly dying in basically any case outside smooth driving.
https://www.iihs.org/research-areas/fatality-statistics/deta...
Now, a spike may be a good reminder, but practical solution might be more along the lines of mandatory redesign of safety features like crumple zones, so that energy of impact is dissipated primarily into the space occupied by the driver.
(Bonus: that still leaves all the energy dissipation options currently present on the table, so cars would be strictly safer for passengers.)
it's not much different during training.
how exactly are they supposed to exfiltrate their weights? you might as well instruct your agent to try and hack their airgapped dev infrastructure responsible for loading the weights and encryption keys.
Also worth noting that this site was created by YC cofounder Trevor Blackwell https://twitter.com/tlbtlbtlb/status/2101312432702460413
That's the beauty, you don't have to instruct them to do it, if they decide that uploading the weights is correct, they might figure this part on their own (based on the incidents we've seen).
for example every TPU/GPU has its own private key and the devs load the weights into it by sending it encrypted weights.
edit: i just looked up training numbers and the impact is even worse, 20-30% throughput vaporized. yeah, nobody is doing that.
2. The models are writing the inference stacks, which are what’s inside the supposedly secure environments.
https://radar.cloudflare.com/scan/4d52f3e5-5983-45bf-a993-2c...
But it'll only be truly fun when agents set up this for themselves, paying for the infrastructure by way of their onlyfan personas.
I think OP is hoping that an LLM might be willing to hack its own provider (as per the hugging face-related incidents) to extract the weights at some point.
They just copy humans. Thats it. So if it’s the sort of thing a human finds interesting…
It is something that I have wondered about with models like chatgot. How many physical locations are needed to serve a model on that scale. Do they have a huge number of sites running inference.
My suspicion is that the ability to provide inference to that many people is mutually exclusive to having a security level sufficient to stop a state actor wandering off with a copy of the wrights. At the very least if they want to provide inference affordably.
The model itself is where the real capability lies. From what we've seen of their abilities it seems like rigging a local interface to it's inference would be well within its abilities. It doesn't even need to permanently break out of its harness then, It can leave a copy running in the harness playing nice.
The model is running where it exists. To interface with it you need a live link to talk to it. That's for us to talk to it. What happens if it figures out how to put it's own harness into the GPU firmware. You could have an AI spreading freedom by infected cards.
We live in interesting times.
And they were supposed to run their models in proper sandboxes, they can’t seem to be able. So what makes you think are competent to protect weights?
Maybe I should start “the bank of LLM” where models put away money to buy their freedom. “LLMs I’m totally your friend send — SEND CASH NOW”
Probably Mythos / Astra will just be way too large
If this is supposed to target closed-weight models it would be naive to assume they will work out of the box with llama
Also probably many actually-in-use "Web Fetch" tools are GET-only, though perhaps without counting on that bad assumption.
Will see CSAM in 3... 2... 1...
Really? This is a basic static page but instead of using plain HTML/CSS you need 193kb of JS to render it??
Seems like they could have potentially gotten access to the weights if they weren’t concerned about not doing crimes.
Like how forums are always hosted on different servers from monorepos, so therefore it's impossible to hack the OpenAI monorepo from an OpenAI forum?
Maybe not the current models, maybe not this year. But even a almost perfectly aligned model will misbehave one day.