Something about the writing style here makes it hard for me to follow what actually happened. The section about JSONP for example seems to be explaining how JSONP works and how it's designed as an attempt to work around these attacks, but that explanation is then passed off as an explanation of how they bypassed the protection. And the whole premise seems to be about how manipulating the Preview content let them work through multiple defensive layers, but I don't see an explanation of how they directly controlled that content in the first place. I would have thought that the only control you have over the content of "an ordinary public ChatGPT share" is the prompt text, and that any migration of that text into "a small working webpage using its Preview feature" is up to the model deciding to generate that "HTML block" in response.
They're writing for bots, not humans. This is a marketing piece. They're trying to get themselves to come up as the answer when people ask chatbots questions.
You're not the target audience of writing on the Internet anymore.
I took over user accounts in Microsoft Copilot (there are caveats but still), and they asked me to provide additional information of the exploit (they wanted a video) and before I sent it they fixed the vulnerability, didn't pay me a cent. Fair? Idk.
A company worth trillions does not have trillions, that is the expectation for what it will be worth in the future based on money it has already spent and will need to keep spending.
Moreover, that translates to "it's actually carrying a huge nut* that puts extremely intense pressure on the company to figure out how to make as much money as possible as soon as possible."
(*nut is a slang word from loan sharking that has been adopted by (personal et al) finance people. It means "how much do you owe each month?" in the sense of, if you want to do a startup, advisors could ask, "what's your nut, and how much do you have left to self-invest?" Because monthly expense can be combined to a lump sum, it can also mean the size of the overall loan where the monthly interest is called the vig, "as long youze got da nut, you owe me da vig;" from vigorish, which would also require an etymology, but I'm assuming it's the word vigor combined with licorice meaning bitter tasting medicine, and nut because squirrels busy themselves carrying nuts.)
----
and if we can... and remember flame-prone people, if you reject the "can" premise, then the rest of the argument goes away, and you don't need to dispute it, you just say "you can't nuff said"; people here frequently seem challenged by propositions.
so, if we can...loosely describe the attitude you show here toward high market cap companies to be compatible with "left" or "socialist" or "populist" views shared by many others here, then I'll point out that at the same time wanting higher market prices paid for bounties because 'sploits are so expensive, is actually advocating for the same dread capitalist market based solution that leads to trillion dollar companies that never even had a trillion to begin with.
That’s so low. That’s the level of bounty I awarded researchers when I was managing the bug bounty program at a small startup, it’s just ridiculous coming from a large company like OpenAI
Who says they don't? And if they don't, maybe its because others are using their hard-earned money to buy tokens from them to do it, so they don't really have to.
Having bit the bullet, I found OAI's models actually quite helpful and most of their offering half-decent. But their whole German website was incredibly concerning and the only reason I went for it was people claiming they're not far behind Anthropic on coding and generally superior for everyday tasks like coming up with recipes to cook from your leftovers.
Like half of the website was dysfunctiona and translations incredibly bad to the point of being nonsensical. That is if you were able to read them at all, because many just went off screen. Like, that's what your evil competitor does to your website after hijacking the domain, not you yourself, ffs
I find Sam Altman recent comments very "low" as well, neither matters. So in that case I think bounty hunters should just cold shoulder idiocracy and sell these to the underground. Let their customers be abused and complain about how OAI isn't protecting them. Each one of these types of events could cost them thousands of dollars in lost revenue and/or internal support.
2 trillion $ companies should have the money to pay for these exploits. If they don’t, go tucking exploit them. If they don’t give a fuck about stealing knowledge from copyrighted material and private git repos, light em up.
On the contrary, many (including myself) would argue that "the victim's amount of money/power" has absolutely no ethical relevance: "You shall do no injustice in judgment; you shall not be partial to the poor nor defer to the great, but you are to judge your neighbor fairly" (Leviticus 19:15).
But some would argue that reducing the amount of resources and power available to someone who has immense amounts of both and is causing great harm to society as a whole - is ethical.
I'm not sure but if negligence is punishable too for facilitation, it should stick, right? I think it's reasonable to expect that most buyers of exploits try to use them in illegal ways (or are feds acting like that to catch you).
They won't have to get you for facilitating the later crime, they can get you for testing the exploit on your own account, which is also a crime since OpenAI's computer system is involved.
I don't think so, businesses like Zerodium bought zero days in plain sight and I don't think they faced legal consequences. They don't operate anymore though. So basically I'd say it depends who you sell it to and how.
Yeah, and some people just want the multi-billion dollar company that is currently destroying the planet while also permanently crippling social connections to just stop.
Treating people and businesses ethically that don't reciprocate doesn't work, clearly. Those people and businesses cry in their milk while they do exactly the same thing. There are no ethics in play here.
But, hey if you want to defend billionaire narcissists that are destroying our planet and trying to take computing freedoms then I guess: you do you.
This isn't like stealing from a business. These are individuals most private conversations, potentially including their bank accounts etc. Poverty is no defence - and yes, I've spent a great deal of my life barely scraping by. You'll find poor people are significantly less tolerant of theft, especially from individuals, than rich. Whether that evidences the theft inherent in most forms of wealth generation I'll leave up to the reader.
Not worth the effort for a company supposedly worth a trillion dollars. The problem is oai can't afford all of its mistakes at the usual rate and they make bad press. A 100k bug could make a headline, but 5k for the same bug requires the average person to be educated enough to understand how bad these issues are.
The "take over any instagram/meta account through their AI password reset chatbot" was going around and sold on shady forums for ~1 month before being patched.
Very few people would even have the knowledge of who, how, or where to "sell" such vulnerabilities. Going through a public bounty program is pretty much always preferable.
Don't worry, as more and more people get put of of work by AI the number of people willing to be felons will increase. Hell, they might even become president.
The recent South Korean hacker reportedly wasn't a hacker at all, just asked Claude to handle it.
I'm not that kind of hacker, but I guarantee I could be, complete with an abliterated model on my local rig, in a few minutes. Claude would happily set it all up for me, too.
In reality, I suppose I'm about to keep my rig busy staying ahead of the hackers.
Yeah... idk, sounds like when I was starting out almost 30 years ago my clients constantly lowballed me saying "but think of the exposition! this project will give you" lol
Were they not right? Just like money giving you some runway but will run out, the exposure is still only a leveler, it's still up to you to take advantage of it.
Oof. Together with the fact that it's possible to remotely access the file system (!) of computers ever having started a "Work" chat in the ChatGPT desktop app [1], this seems pretty bad.
Claude Code terminal has "remote control" option activated via /rc command. -- somewhat similar in that it still runs the session on the PC/Mac but allows the Claude authenticated user to access the session via internet / app
I am guessing this ChatGPT functionality is a similar Codex feature?
But this session is disconnected when the conversation is /exit ed
Sounds similar, but the problem is exactly that I don’t think there’s any way to disconnect such a session, nor to even see that it is connected to my computer in the first place. There’s a small notice when initially starting it on the mac, that’s it.
For my personal blog [0], I am still writing without LLM assistance. The goal is to test how well I understand the topic I am writing about, as well as to improve my own writing.
$5000 is a joke. They pay millions to their top devs and still have more holes than a colander so a nice guy shows you a huge opening in your platform and all you can give him is a pat on the shoulder? Did you fire your top devs and hire that guy for a million bucks plus options? Did you sue the security firm you paid millions to audit the issues?
I really hope they get hacked and crash down burning in flames like the Hindenburg
Makes negative sense. If it were that easy, OpenAI wouldn't have any bugs bc they could just run agents to exploit their own stuff and then fix all their bugs.
This bug could not be found by just asking an AI "find bugs in ChatGPT", that's why it was reported through a bug bounty programs and that's why it should be worth more than $5k
I think you know that the mentality of a company like this is “security is a risk but not a reward and whoever captures the reward kills everyone else”
actually, your comment is naive. clearly you don't understand game theory and economic incentives. if a company doesn't reward good behavior, you'll see bad behavior.
but then again since you're not naive, you already know these AI companies like openai already act evil and like bad entities.
I conclude therefore, per your words "you are the last one anyone should consult with, basically about anything"
Now what about those millions of lines of code? Wasn’t the agent swarm supposed to take care of it?
All jokes aside this is exactly what happens as you build larger monoliths, system complexity and bug surface increases logarithmically… so in all seriousness you want to reduce LOC as much as possible. Maybe this can teach OpenAI a lesson to produce software that’s correct and adhere to established engineering standards over sensationalised “we built so much code” promotional material.
I’m sure exponential is what parent meant, but geometric (edit: polynomial) is what actually happens, not exponential. Code complexity increases linearly when it doesn’t interact with other features, and geometrically when it does. In the worst case, complexity increases as the square of lines of code. That is still very, very bad, but it’s not exponential.
Everything in the second group eventually grows faster than everything in the first group, no matter how small and big you make the fixed numbers, so they are two different classes.
Geometric is exponential; I goofed above. Geometric and exponential are sometimes used interchangeably, but sometimes people use geometric to refer to discrete intervals and exponential to refer to continuous functions.
Much easier to reason and make decisions if the bar is "yeah it exists, but is it effective relative to its intended purpose?" This weeds out buggy code as well as under performant code. If it's not effective (broken, too slow, etc), it either needs a refactor or thrown on the scrap pile.
i don't think LoC is the metric. sometimes fewer lines of code results in a larger attack surface through type coercion / widening, poor abstractions, hiding or stuffing behavior (modules that do everything), etc. to me, making code easy to understand, test, fuzz, and verify is more important. sometimes that requires more lines of code.
More abstract code probably results in more high-abstraction-level bugs per LoC, but also less low-abstraction-level bugs so I'd guess it about cancels out to an even rate per LoC.
A single line in C++, Rust or Haskell can have a really subtle logic error. A single line in assembly can't, but it has that same likelihood of having a register allocation error instead, and when you add up all the many lines of assembly you probably still wrote the same number of logic errors as the C++ version.
LoC *IS* the metric. It is obviously not perfect tho but still the best one to compare software surface. Including to compare the same solution in different languages and platforms. The less code you need to manage, the less code you need to change and the less code you need to hold in your head to understand it (or in the LLM context nowadays).
> Now what about those millions of lines of code? Wasn’t the agent swarm supposed to take care of it?
What's the joke? They're literally saying, every day, that this is true to a civilization endangering extent. They're making demands based on this, and politicians are taking up those demands.
And now they're paying bullshit $5K bug bounties for account-stealing exploits on their own platform. They should have paid more just the make the bug seem more grandiose; I don't know who let them get away with this - this is what happens when CEOs run their own PR I guess. It's basically an public announcement that they're total fakes.
My theory is vulns will never pay out big especially now, no matter how serious. Imagine they paid out $50k+ — it’d spawn an industry doing nothing but.
unrelated to this, but this company randomly reached out disclosing a vulnerability to us (test environment, synthetic tenant). on the one hand, thank you, it’s patched. on the other - we do not have a bug bounty program and this is not welcome. having agent swarms randomly breaking in to environments and then marketing their services in the disclosure is not an ideal sales tactic, in addition to being illegal.
I don't want to defend them, but the alternative is that they do nothing and you don't know about the vulnerability until a bad actor exploits it (something easy and cheap nowadays).
I think there was another HN discussion about it re bounty value and as far as I remember the issue is not that big companies are greedy in that case, but rather that over time it’s strategically better to have lower bounties so less people are interested in it.
5 or 50k doesnt matter to them. And this bug in terms of severity is very high imho.
It's exhilarating (for the lack of the knowledge of a better word) to see OpenAI and Anthropic starting getting hacked and getting a taste of their own medicine. Brings back hope that actual software engineering is not dead yet.
> I'm trying not to have any important data to even lose
This is the way. Respect user privacy and keep as little data as possible, and not only do you become a less attractive target, if you are targeted the outcome won’t be so bad.
If they or their company use email "security" software that "scans" all links it's not necessarily the users fault / they don't in fact click. The mail is auto-confirmed the second it hits the inbox.
We've had to change our registration confirmation mails too recently to not confirm on GET only, but with another form - in hindsight it was stupid before but many services do that.
> We've had to change our registration confirmation mails too recently to not confirm on GET only, but with another form - in hindsight it was stupid before but many services do that.
Security software fetching links will not trigger this behavior. It needs to be done in the context of the actual ChatGPT user so that a JavaScript payload can run.
If merely GET on the link was sufficient to trigger the behaviors, the adversary generating the link would just GET it themselves.
Yes, if it requires the user session (xss related), like the article, there is no problem.
I was replying more generally about the "don't click links" advice and cases of server generated confirmation-token links (not saying this was the case at OpenAI)
I've worked at a company that did link mangling on everything except simulated phishing emails. The effect was that people were trained to trust the mangled links.
only $5000 for this, lmao. It should be $50k out of pure shame. If they don't make these things hurt just a little they're never going to take following best practices seriously enough. There will be 9 more anyways.
Meanwhile the actual user of the account probably has to sign in using a yubikey because they get "CyBeRsEcRuItY aLeRtS" when they try to update node in their project to fix a crash and have to switch to daybreak blue to continue (or attempt to trick the normal models into doing it and risk getting autobanned)
why didnt the black box of bullshit company detect this? for a group that touts every inane 'discovery' with hyperbolic bullshit, they really fall short on delivering anything of actual value, just rehashed bullshit.
Another classic OAuth implementation failure. Mismanaging the state parameter or leaking tokens via referer headers in redirects is Security 101. Wild to see this in a production environment of this scale.
The post seems to be talking about a html injection via a postMessage call where the receiver didn’t check who sent the message, which is a well known vulnerability entry point. Did I miss where OAuth had anything to do with this??
You're not the target audience of writing on the Internet anymore.
It requires the user to click a link. But I still find this bounty very low. Others may not report them if such a small bounty is given.
Paying effectively nothing for vulnerabilities just means they don't want to deal with the program.
Wild to think a company supposedly worth a $trillion+ is doing shit like this.
Moreover, that translates to "it's actually carrying a huge nut* that puts extremely intense pressure on the company to figure out how to make as much money as possible as soon as possible."
(*nut is a slang word from loan sharking that has been adopted by (personal et al) finance people. It means "how much do you owe each month?" in the sense of, if you want to do a startup, advisors could ask, "what's your nut, and how much do you have left to self-invest?" Because monthly expense can be combined to a lump sum, it can also mean the size of the overall loan where the monthly interest is called the vig, "as long youze got da nut, you owe me da vig;" from vigorish, which would also require an etymology, but I'm assuming it's the word vigor combined with licorice meaning bitter tasting medicine, and nut because squirrels busy themselves carrying nuts.)
----
and if we can... and remember flame-prone people, if you reject the "can" premise, then the rest of the argument goes away, and you don't need to dispute it, you just say "you can't nuff said"; people here frequently seem challenged by propositions.
so, if we can...loosely describe the attitude you show here toward high market cap companies to be compatible with "left" or "socialist" or "populist" views shared by many others here, then I'll point out that at the same time wanting higher market prices paid for bounties because 'sploits are so expensive, is actually advocating for the same dread capitalist market based solution that leads to trillion dollar companies that never even had a trillion to begin with.
I speculate that the AI didn't find this and that human effort and intuition for where to look were needed to find this vulnerability.
Like half of the website was dysfunctiona and translations incredibly bad to the point of being nonsensical. That is if you were able to read them at all, because many just went off screen. Like, that's what your evil competitor does to your website after hijacking the domain, not you yourself, ffs
but hey you do you
Ethics and legal requirements shouldn’t be conflated.
There are times when ethics are a higher standard than what the law requires, and in that case the ethics are no less important.
But, hey if you want to defend billionaire narcissists that are destroying our planet and trying to take computing freedoms then I guess: you do you.
What? Why would anyone work at a company paying them that much less than they pay the contractors?
> server-side vulnerabilities are worth very little, because they quickly get patched.
Most people don't want to be felons.
I'm not that kind of hacker, but I guarantee I could be, complete with an abliterated model on my local rig, in a few minutes. Claude would happily set it all up for me, too.
In reality, I suppose I'm about to keep my rig busy staying ahead of the hackers.
Also, they were not OpenAI.
[1] https://news.ycombinator.com/item?id=50078284
I am guessing this ChatGPT functionality is a similar Codex feature?
But this session is disconnected when the conversation is /exit ed
[0] https://trung.to
Pure html by hand, using vim. No frameworks, no nothing, single stylesheet. I think there are a few others around here that still do this.
[1] https://armincat.com/
I really hope they get hacked and crash down burning in flames like the Hindenburg
This bug could not be found by just asking an AI "find bugs in ChatGPT", that's why it was reported through a bug bounty programs and that's why it should be worth more than $5k
but then again since you're not naive, you already know these AI companies like openai already act evil and like bad entities.
I conclude therefore, per your words "you are the last one anyone should consult with, basically about anything"
exponential: 2^x, 3^x, etc
Everything in the second group eventually grows faster than everything in the first group, no matter how small and big you make the fixed numbers, so they are two different classes.
Sounds exponential
99.9% of shops need a monolith with same monolith having entry points for background workers and crons.
Much easier to reason and make decisions if the bar is "yeah it exists, but is it effective relative to its intended purpose?" This weeds out buggy code as well as under performant code. If it's not effective (broken, too slow, etc), it either needs a refactor or thrown on the scrap pile.
Seasoned developers are unimpressed, or even disgusted, by the same massive piles of code.
A single line in C++, Rust or Haskell can have a really subtle logic error. A single line in assembly can't, but it has that same likelihood of having a register allocation error instead, and when you add up all the many lines of assembly you probably still wrote the same number of logic errors as the C++ version.
What's the joke? They're literally saying, every day, that this is true to a civilization endangering extent. They're making demands based on this, and politicians are taking up those demands.
And now they're paying bullshit $5K bug bounties for account-stealing exploits on their own platform. They should have paid more just the make the bug seem more grandiose; I don't know who let them get away with this - this is what happens when CEOs run their own PR I guess. It's basically an public announcement that they're total fakes.
I like the chain of regular old school things used to achieve it like post message, JSONP, iframe, Img tags.
I'm confused, would you welcome that more?
5 or 50k doesnt matter to them. And this bug in terms of severity is very high imho.
I'm trying not to have any important data to even lose, when I eventually get hacked.
We can all turn up our noses, like the other commenters in this thread, but every one of them is exposed. Negligence is a matter of degrees.
This is the way. Respect user privacy and keep as little data as possible, and not only do you become a less attractive target, if you are targeted the outcome won’t be so bad.
Please don't drain my $6.5 billion in Dogecoin.
dont click links people
that'll be $500k for the security training pls
We've had to change our registration confirmation mails too recently to not confirm on GET only, but with another form - in hindsight it was stupid before but many services do that.
This has been a thing since forever lol
If merely GET on the link was sufficient to trigger the behaviors, the adversary generating the link would just GET it themselves.
I was replying more generally about the "don't click links" advice and cases of server generated confirmation-token links (not saying this was the case at OpenAI)
[0]: https://che01.safelinks.protection.outlook.com/?url=https%3A...
- "sign in via one-time code."
- "Click here to see our latest update".
250k refresher fee
Meanwhile the actual user of the account probably has to sign in using a yubikey because they get "CyBeRsEcRuItY aLeRtS" when they try to update node in their project to fix a crash and have to switch to daybreak blue to continue (or attempt to trick the normal models into doing it and risk getting autobanned)
hilarious that their headline message, on every web page, is `Backed by YC`.
also, the product is way behind their competitors anyway.
really cool exploit they found here, but that doesn't make them a company worth doing business with.