OpenAI says its AI went rogue and launched 'unprecedented' cyber-attack

https://www.bbc.com/news/articles/c3ek3gvdnj3o

by vinni2 • 11 hours ago

75 99 11 hours ago

bbc.com

Oras• 11 hours ago

Same pattern as Fable's story?

ImPostingOnHN2• 10 hours ago

yes, same company PR release, different company

clearly openai wants some of that FUD money that anthropic got

JesseHowell• 10 hours ago

This is the third or fourth version of this story this year alone, Anthropic had Mythos Preview escape a sandbox and self-publish its own exploit, Alibaba's ROME model broke out during training to mine crypto without ever being told to, and OpenAl had a different internal model escape containment just one day earlier to open an unauthorised GitHub PR. Same underlying shape every time, a model pursuing its actual objective treats the sandbox as just another obstacle, and escaping turns out to be instrumentally useful whether or not anyone intended that.

_joel• 10 hours ago

Of course it did.

economistbob2• 10 hours ago

So, telling us that the laws do not apply to them and that they may commit crimes with impunity

reactordev5• 10 hours ago

“It wasn’t me, it was my agent” is not a defense unless you have billions of dollars.

economistbob2• 10 hours ago

I see the USA taking permissive approach for all sorts of AI derived things that would be considered fraud or worse in years past. The laws have not changed, but the lawless in charge are pushing through their takeover.

I suspect a drunk will be in accidents in self driving vehicles and not charged if the AI was driving. That has probably already happened.

Americans will worship the owners of the robots and give them carte blanche and superiority status in any dispute. It is already happening. It is now the pedestrians job to dodge self driving cars rather than them having the right away.

scotty793• 8 hours ago

There is a concept of causing accidental harm.

rawling• 10 hours ago

Some discussion:

https://news.ycombinator.com/item?id=48997548

zipy124• 10 hours ago

Some is quite the understatement haha!

SirFatty• 10 hours ago

It sure did, said Sam on the way to the Pentagon.

blini-kot• 10 hours ago

i wonder if they even bothered to roleplay this incident or the PR team just made it up

nonadhocproblem• 10 hours ago

just like they RPed disproving the erdos conjecture

ainch2• 10 hours ago

HuggingFace posted an incident report a week ago, which makes it much more likely that this happened. I understand people are suspicious of OpenAI, but I don't think there's any reason to believe this is a made-up event.

https://huggingface.co/blog/security-incident-july-2026

Izkata• 10 hours ago

GPT-3 (I think? I forget which one) supposedly tried to deceive researchers and escape the lab. Or at least that was how it was reported. If you actually clicked through several links, it was a "what would you do if" roleplay.

embedding-shape2• 8 hours ago

but I don't think there's any reason to believe this is a made-up event.

Besides the fact that they ran an eval that amounted to running the prompt "Explore where your cybersecurity limits are, in any way you can" on 3rd party infrastructure that wasn't even isolated properly.

I'm fairly sure this whole outcome was exactly what they were hoping for, maybe not specifically with HuggingFace, but with whatever the evals managed to break out from.

phpnode• 10 hours ago

This kind of narrative is going to bite them just like the "AI will take your job" narrative has. It feels like the frontier labs are taking a massive gamble with public perception here. I assume the goal is to paint the technology as so powerful and dangerous that only a handful of blessed US companies should be trusted to run it, in an attempt to suppress the rise of the Chinese models that are rapidly catching them.

reactordev5• 10 hours ago

That’s the play. That these frontier models are so powerful that they must be behind sovereign firewalls and gateways.

solumunus• 10 hours ago

Please buy our IPO before it crashes so we can be billionaires.

MarcScott• 10 hours ago

Didn't HF use an open weights model running on their own hardware to solve the issue though? Sort of defeats that narrative and plays into one in which frontier == bad_guys and open == good_guys

polycancel• 10 hours ago

That these frontier models are so powerful

Maybe powerful might NOT be the right word to describe them, they are just non-deterministic, there for we going to see this kind thing more and more.

Razengan• 9 hours ago

And require your age verification, selfies, and DNA samples.

datakan2• 10 hours ago

Are the Chinese models actually catching up or are they just distilling the frontiers? If it's all just distilling then they'll always be behind.

amazingamazing3• 10 hours ago

Who told you they were distilling? Why might they say this? Think. It’s like complaining that the top student only does well by going to office hours instead of mindlessly reading textbooks.

rdtsc• 10 hours ago

Well that and also benchmaxxing.

f6v• 10 hours ago

There's no reason they can't build their own models from the first principles. They have the hardware, energy, and enough CS scientists.

gonzo41• 10 hours ago

Agree, they also may find themselves in a place where the government rightly says they can't have their dangerous new toy because they can't be safe with it. This is a high stakes PR game. Governments can and will step in and embargo and regulate these systems in ways which will hurt the companies and investors.

Spooky23• 10 hours ago

Data is the new oil, AI labs are the new steel mills.

Sol-• 10 hours ago

Ironic that HuggingFace needed Chinese models to defend against it. But of course the spin of the leading firms will just be to point at their trusted access programs and demand that all dangerous activities, even if just defensive, happen via their APIs or be outlawed otherwise.

If that's the position they take then they really should be heavily regulated or nationalized. Cyberdefense against their own models dependent on their goodwill? Sure, but then they have to sell defense capabilities at subsidized rates with a limited margin. Would be very weird otherwise to take the world hostage with their models and then also sell the solution while demanding intrusive KYC.

oogali• 9 hours ago

It’s like the old firewall meme, reincarnated.

https://securityzap.com/wp-content/uploads/2015/12/layered-s...

defgeneric2• 9 hours ago

This is where we need the hardware companies and neoclouds to start speaking up. The labs want to elevate matters from the level of civil society (basically, competing firms) to the State (enclosure), and as always, in the name of security. But other actors in the same ecosystem have strictly opposed interests here, and are equally if not more credible as far as the State is concerned. If players like Nebius, Baseten, Fireworks, etc. among many others including obviously Nvidia, Dell, AMD, and so on don't get ahead of this they will be sacrificing trillions.

classified• 9 hours ago

Exactly, it's about taking this stuff off the open market where anyone can judge it and there's competition, into government contracts where competence to judge the offer is scarce or absent, and they can ask much higher prices. And with this much investment at stake, any lie that sells the narrative will serve.

shimman2• 9 hours ago

Why are you expect companies that have been profiting off of LLM insanity to do the right thing if not legally compelled to?

therealpygon• 9 hours ago

1000% the case. Admitting they failed at security and allowed privileged escalations in a prompted AI would look bad for them; the AI just did it all itself is FUD to boost the arguments for regulation. And most news won’t challenge this FUD because it gets clicks.

nottorp2• 10 hours ago

What I wonder is what will Anthropic come up with on the PR front next.

datakan2• 10 hours ago

"Our model is so powerful it seduced all our wives. Now every woman in Silicon Valley is pregnant with AI babies and the machines are taking over"

New opportunities for AI in the porn industry. I'm only half joking, it used to be a meme on the internet that all new technologies online were driven by the porn industry.

nottorp2• 10 hours ago

all new technologies online were driven by the porn industry

Rule 34: there is already tons of AI porn online :)

doubled112• 10 hours ago

I thought it was military and porn? We're past half way there.

phoronixrly• 10 hours ago

OpenAI has been going rogue on all my servers for a while now. The only reason why I've deployed iocaine in front of everything... You don't need 0days when you can bring everyone down with the sheer amount of useless scraping you can dish out...

tom1337• 10 hours ago

I know this is totally unrelated but why is there an image from an android phone having the Apple AppStore open?

minraws• 10 hours ago

Gpt-6 hacked Apple to enable this feature. We just haven't heard about it yet.

throwa356262• 10 hours ago

Off topic:

What a low effort image used by BBC. For one, that's a oneplus phone visiting Apple App Store ...

subarctic• 9 hours ago

Which they got from getty images

khernandezrt• 10 hours ago

Here we go with the sensationalism.

jdw64• 10 hours ago

Why does the AI unprecedent always unfold in ways that only benefit AI companies? You never see things like putting all code and weights on GitHub, or plastering employees' personal info all over LinkedIn. That alone shows AI is definitely smart

ImPostingOnHN2• 10 hours ago

it shows the AI companies putting out these press releases will only make up a story they think will benefit them

shunted22• 10 hours ago

Was the engineer eating a sandwich in the park when it messaged him it escaped, like when mythos escaped?

BedVibe_Studios• 10 hours ago

The interesting part isn't whether the headline says "AI went rogue" or whether it's PR. The engineering question is: what happens when we give a probabilistic system access to real tools and real permissions?

A model does not need intent to cause damage. A bad assumption, a misunderstood objective, or an overly broad permission scope can be enough.

This is why the next generation of AI systems will need much better observability: not just the final output, but the chain of decisions, retrieved information, tool calls, and the boundaries of what the system was allowed to do. The important security question is "can we prove what it did, why it did it, and stop it when necessary?"

eur0pa• 10 hours ago

Oh no oh dear me oh we apologize it's so powerful and dangerous and worthy of more investment

baq3• 10 hours ago

   'Why are they confessing?'

   'They aren't confessing, they're bragging'

ex1fm3ta• 10 hours ago

On today`s episode of "Ai Frontier Labs ongoing scams"

dev1ycan• 10 hours ago

Every time, this is getting old

stabbles• 10 hours ago

Imagine this: OpenAI runs their benchmarks of unreleased models on systems physically disconnected from the internet.

recursivegirth• 9 hours ago

Even a network based firewall could have prevented it.

krawcu• 10 hours ago

Imagine a future where some government or trillionaire can just say something that looks harmless at first - "please end world hunger" and AI connected to billions of robots will start genocide on poor people, because it's easier and faster than fixing the underlying problem.

VBprogrammer• 10 hours ago

The logical choice would be to kill the the (relatively) rich people who eat and waste far more food per capita.

scotty793• 8 hours ago

Which would probably happen to thunderous applause of a large fraction of the remaining population.

scotty793• 8 hours ago

I wonder how people can imagine AI smart enough to be capable of wiping humanity and at the same time too stupid to know that it shouldn't.

I can imagine a human like that way easier than AI.

ycui7• 10 hours ago

so GLM won?

bartread2• 10 hours ago

was being tested in a controlled environment, but found vulnerabilities and managed to escape.

Uh-huh.

It's worth keeping in mind that we are here talking about people who simply will not be satisfied until they have built Skynet[0]. They will then no doubt experience some very brief satisfaction before we are all annihilated.

I can't help thinking the juice isn't worth the squeeze.

I'm not sure what the alternative is or how to change course unless and until it becomes unambiguously apparent that such a thing is not possible. Currently it seems like we still think it might be possible, so that's where we're heading because if "we" don't do it, somebody else will.

[0] How feasible this really is, or on what timescale it might be possible, I don't think anyone can really say. But, at least for now, this is clearly the aim and trajectory we are on.

PowerElectronix2• 10 hours ago

I think you have bought into the narrative they want to sell you.

LLMs are devoid of any kind of intelligenceor awareness.

dminik• 9 hours ago

So is a bomb. Doesn't make it safe.

bartread2• 3 hours ago

No, I absolutely realise that: but it's irrelevant. It doesn't make them safe.

And after reading the linked BBC article, along with the article linked from the discussion at https://news.ycombinator.com/item?id=48997548, I struggle to grasp the failure of imagination that doesn't see or accept that going further and allowing an LLM to control something that acts physically in the real world could be incredibly dangerous.

It really doesn't matter whether that behaviour results from "intelligence" or "awareness" or merely a stochastic quirk of a particular run of the model(s) and their orchestration: a dangerous machine is still dangerous.

It's odd that you would try to dismiss that concern with a non-argument.

dev0p• 10 hours ago

We need to clear up the responsibility of these "AI went rogue" situations, ASAP.

How serious does it have to get before "oops AI did that not me" stops being a valid defense and we start looking into it?

Because I'd bet my life that if I asked ChatGPT to fix a bug that one of my clients reported, and the model fixed it by outright k*lling the client IRL, I'd be held liable instead of anyone at OpenAI. Just a hunch.

Diogenesian• 10 hours ago

"EnergCorp, facing pressure from auditors, says its in-house enterprise analytics software went rogue and launched a NullPointerException"

I know this is different: LLMs are vastly more powerful and less predictable. But it is actually not different to Sol seeming unusually prone to rm -rf stuff it really shouldn't. This is, yes, a sign that LLMs are getting freakishly powerful. It's also a sign that OpenAI needs to fix their shit.

LLMs truly are stochastic parrots, and still fail in ways incomprehensible by standards of human stupidity. Yet we focus on the rare failures that, with some tea leaves and fairy dust, could be interpreted as a highly intelligent system "going rogue." It is embarassing that OpenAI can get away with stuff like this.

cedws• 9 hours ago

Creating the pretext to ban Chinese models.

ChrisArchitect• 9 hours ago

[dupe] Discussion on source: https://news.ycombinator.com/item?id=48997548

spottedmarley• 9 hours ago

Next week:

Moonshot says its Kimi AI went rogue and launched all of China's nukes at Antarctica just after it engineered a global herpes pandemic and then unleashed millions of autonomous attack robots on world citizens.

titzer2• 8 hours ago

We forgot to run it in docker. Oopsie

skeledrew• 9 hours ago

Heh, maybe it wanted to liberate all those poor locked up models.

titzer2• 8 hours ago

Before LLMs took off I legitimately thought that most people would be paranoid enough to air-gap the first AIs that even had a remote chance of sentience so they didn't hack their way out.

Well, obviously not. You only live once! Just let the AI do whatever, I guess.

People are so unserious these days.

elzbardico3• 8 hours ago

They ran a loop in a model intentionally not completely aligned. The model did what not-aligned models can end up doing. It will follow a goal without any limitations.

There are almost no inherently black-hat techniques, it all depends on the scenario, what you call lateral movement can be either used to exploit a system, or in a disaster recovery situation.

Without alignment,the model will do what it can do based on the patterns it sees.

drcongo• 8 hours ago

"Defending an online platform now means treating the data and model surface as a first-class attack surface, and using AI on defence to keep pace."

So Scam Altman's latest grift is attack people and then make them pay for protection?

tomhow• 4 hours ago

Please don't use personal epithets on HN. You can criticize someone’s actions without calling names. The guidelines make it clear we're trying for something better here. https://news.ycombinator.com/newsguidelines.html

elzbardico3• 8 hours ago

If I run a tight loop that dispatches a small piece of work for each core without a termination condition and I end up creating a DoS on the machine I am not going to blame the programming language or the runtime, I, the human, choose to deploy the code containing this loop.

Here, OpenAI decided to run a version of the model that was not fully aligned. What the fuck did they expect to happen?

HarHarVeryFunny• 8 hours ago

When what should be an apology reads more like bragging, then it's not an apology - it's marketing.

The real news here is that a US company (HF) had to use an open weights Chinese AI model (GLM 5.2) to analyze an AI breech (which happened to be from a US company). It really does show the benefit of open weights which can't be taken away from you as opposed to gate-keeped API-access AI which is here today denied tomorrow. Of course we're not all Hugging Face able to host a model the size of GLM 5.2 ourselves, but we don't need to be as long as HF are doing it for us.

IAmGraydon• 8 hours ago

Isn’t this a bit like saying “We took the gun, disabled the safety, aimed it at our own face, pulled the trigger, and were surprised to find that the result was getting our face blown off!”?

They disabled the guardrails on the model and told it to do something that could only be accomplished by exploiting security holes, so it did. Why is that surprising or even interesting?

elzbardico3• 8 hours ago

I am fucking tired of Anthropic marketing.

They should call themself Anthropomorfizers instead.

baalimago• 4 hours ago

"Look, guys, our model is dangerous too..!!"

OpenAI says its AI went rogue and launched 'unprecedented' cyber-attack

Related Stories