The commit $1 billion in tokens to "defenders". $1 billion directly to QA for open source projects [1] which would write 100% code coverage, run ASAN and fuzzers would be more effective.
But of course there would be no marketing.
[1] From whom you stole everything in the first place.
I'm as frustrated as the next open source maintainer, but "stole" goes too far. We gave that code away with pride, knowing the world would be a better place.
Few works are released directly to the public domain. It is my understanding that the public domain doesn't even exist in all legal jurisdictions.
Most people license out the software under specific requirements and prohibitions. BSD requires attribution. GPL variants require adherence to copyleft when distributing derivative works. A little under half of all github repos don't even have a license, which means that though the source is published, no one has any license under which to copy it or derive new works from it.
So far as I have seen, whether aggregating and munging these works into abstract LLM weights constitutes fair use is still legally unclear. If not deemed fair use, it is almost certainly a massive violation of copyright.
You can read code, rewrite the functionality yourself, and sell it, and not be at odds with copyright. Open source does NOT protect against reimplementation. See Java lawsuits.
If you don't want your code to be used as possible inspiration, then you are not actually for open source, and are using the wrong license. The licenses are legal bindings, not moral bindings.
"Don't use any concept found within this code" would be a license that doesn't really fit with any current copyright law, and it wouldn't allow any software to exist, except that made in the 70's.
>You can read code, rewrite the functionality yourself, and sell it, and not be at odds with copyright
Traditional reimplementation often separates the roles of investigation and implementation specifically to avoid assertions of copyright infringement. WINE, for example, does not accept patches from anyone with prior experience in the Windows code base.
By training and generating from the same the LLM, as one necessarily must do as LLMs are built from training, you inherently violate this distinction.
>If you don't want your code to be used as possible inspiration
Inspiration is a human affect. Training an LLM and generating tokens therewith are mechanical processes.
Courts have not yet decided whether LLM data mogrification truly constitutes fair use.
>then you are not actually for open source, and are using the wrong license. The licenses are legal bindings, not moral bindings
It is not for you to decide what rights someone else wishes to attach to code they release. There are many open source licenses for those wishing to grant various sets of rights. Your reference to morality seems more in line with the GPL specifically, as the Free Software Foundation has a moral stance on freedom in computing not necessarily shared by other open source licenses. 'Free Software' is the term taking a moral software stance.
>"Don't use any concept found within this code" would be a license that doesn't really fit with any current copyright law
Yes. That would be a patent.
>and it wouldn't allow any software to exist, except that made in the 70's.
Patents only last 20 years. They have frequently delayed the spread of certain algorithms and data file formats under their restrictions. This has frequently lead to open source, and proprietary for that matter, being forced to invent their own methods to work around the prohibited ones.
That's as noble as throwing breads at the crowd while being perfectly aware that there lurks a number of armed henchmmen of "noblety" in it, who won't miss the opportunity to amass those breads to resell them.
Or at least just childishly naive for failing to notice that beforehand.
Speak for yourself. I don't see every model being released under GPL after having read the entire source code as a step of creating their binary artifacts (model files).
The point of the open source I work on is to contribute to a commons, not to have it stolen so that it can be used to destroy those commons, used for the profit of techbros, and used to justify the murder of children.
To be further pedantic, is that really stealing? They are disregarding copyright but was anything taken from anybody? I thought that was the old argument for piracy, ie, nobody was deprived of anything.
It’s marketing, but in a global good will sense of “look how much OpenAI cares about secure open source.”
This works for established profitable brands that want to build / maintain a good public image.
This isn’t the marketing needed by an unprofitable startup that wants to show hyper growth before its IPO. They need the “use our product or get pwned” kind of marketing.
"Daybreak enables vetted enterprise customers and cybersecurity practitioners to use our most capable models for dual-use cybersecurity work."
They explicitly give out access to those who will make a big announcement of bugs found. OpenAI does not want projects to fix bugs themselves so it can have more bugs and better marketing.
The first sentence of "Coordinating a global response" is "Getting a fix into use can mean adapting it for older releases"... but I guess they mean "only supported releases", as the fixes only go back as far as 8.2, which still has security support until the end of the year, but nothing for 8.1 or 8.0, and in the case of PHP there are still plenty of PHP sites on 7.x and even 5.6.
I wouldn't expect the PHP team to provide patches for versions for which security support has ended. But this seems like the kind of task that AI is especially well suited for-- rework the patch for a release that is no longer supported but still widely used. It would've been nice for people who have much older sites to be able to patch their PHP rather than have to upgrade PHP as well to get the fix.
It seems like OpenAI are trying to frame this as if AI did all the work here, but it's not really true given what Calif say (https://calif.io/php2shell/):
> How did we find this?
> No special harness, or access to models that weren't generally available to everyone were used in the initial bug discovery. I focused on something I had a background in, knew its vulnerability history well, and held a theory that specific bugs lurk within the codebase. This shows that "know your target" is still as applicable as ever, even in the LLM dominated VR age. No model is simply capable of unearthing "all bugs" with simple prompts. At least for now.
> There are some practical research implications here. LLMs may have ingested public historical vuln data, but in their default variant analysis tends to move from the newest and most prominent bugs toward nearby older code. They do not reliably begin with obscure twenty-year-old exploitation primitives. Keeping your own mental catalog of a target's vulnerability history lets you prompt that direction explicitly.
They literally open by saying Calif used GPT 5.6 Sol, which is a widely available model. Yes, the person driving did more than "find all the bugs", but nowhere does OpenAI imply otherwise.
> It highlights how AI can uncover security flaws with internet-scale impact
> To our knowledge, frontier AI found these vulnerabilities before attackers did
I don't know who/what Calif is and there's no link to them, so if I were only reading this article I would definitely get the impression that "AI did all the work here".
Pure marketing, it's getting increasingly hard to separate actual technical feats of agents from marketing pretentious declarations. Such a terrible environment to navigate right now.
Damned if you do, damned if you don't. Do you prefer they not engage in helping finding vulns at all? Do you not think there are parts of the company interested in security and doing it for the sake of that?
I feel like taking every action these companies take as part of some larger strategy is kind of conspiracy-minded.
This is a false binary. There are a range of other options. They could voluntarily log every action with a clear warning that users are financially and criminally liable for use of AI. The company could convert to a partnership like law firms and investment banks used to be so that the executives are personally liable for harm their products cause. Hard to imagine they would be as flippant with risk if they would lose everything in the event of a
Cyber attack. We could set up a parallel internet for AI that is physically separate from human internet and limit ai to text messages with all tool execution on the sandboxed internet.
I could give 15 more examples of things that could be done other than “make the world use our tools to patch software”.
Disclaimer: I work as a sec engineer and part of my job includes finding vulnerabilities.
I think the effort is misplaced. We should be exploring ways to make our systems more resilient and I would expect those companies* to focus on that front. Vulnerability research is important, and we should aim to remove as many bugs as we can from our codebases but the pace of software development is increasing and so is the introduction of new bugs.
* I would also expect an increase on relevant research regarding novel paradigms of securing systems.
The conspiracy theory is suggesting this is part of a larger strategy. I'm just saying it's a big enough company with a wide enough range of beliefs that I find it pretty likely that the direct motives for this weren't to enable regulatory capture.
But they could be, that's why it's a conspiracy theory.
I don’t think it’s marketing but the fact that all the labs don’t immediately agree to stop further release until security is more established shows the profit motive is inappropriate for this industry and needs to be removed from them. You don’t get to threaten humanity while cashing in stock to buy a San Francisco mansion.
You’re right overall, but keep in mind that part of the way “US imperialism” works has been to export its corporate style to other countries. This means the infection goes far beyond the US alone.
keep in mind, the point of most views is a toddler who lies about stealing a cooking isn't corrupt, because they're a child and can't do anything.
The same view point is used for racism, dei, and the rest: it doesn't matter if the poor beggar on the corner hates white people. They don't have any means or methods to act on it.
> Calif identified a critical flaw using GPT‑5.6 Sol as part of our Patch the Planet (opens in a new window) initiative
> 19–24 August 2026: Wiz Research independently collides with the same bug family using Wiz Atlas, Gemini 4 Argon, and GPT-5.6 Sol, reporting its variants on 24 August.
I again ask how they are lying? I get if people hate these companies, but blatantly disregarding what is said helps nobody.
> OpenAI are trying to frame this as if AI did all the work here
CEOs of all big corporations are constantly lying and misleading the public, investors and shareholders. That is a crime. A shame that it is not enforced anymore.
> How can we help identify and fix vulnerabilities in downstream packages while upstream fixes are pending, or backport fixes to older versions that no longer receive security updates?
Is the company selling a magical code generation machine really asking how it can backport a fix to an older version of the same software?
Unless the magical code generation machine can update packages that are no longer being updated, this isn't really a problem that can be solved with magical code generation.
Backporting the fix is much lower effort in total. It's just a few dozen minor PHP releases that are affected, but millions of websites still running on older PHP versions.
people running "older versions that no longer receive security updates" are unlikely to update to some new patch-stream from this tooling, because they're already not maintaining their infrastructure.
The LLMs can't accept, sign, package, and release fixes to deprecated software on behalf of unaffiliated maintainers. If the software is EOL, it's going to stay that way.
no. I think they mean how to navigate the non technical organizational challenge, not how to technically fix the thing
it’s a quite simple matter to fix a bug once it’s discovered, but how to deploy and where to deploy it when it affects a supply chain?
here’s an example: you can submit a vuln at a downstream dep and they can tell you it should be patched upstream. what should you do? open an issue upstream? maybe, probably not, probably the solution is to contact all downstream deps in order to get them to migrate dep or patch, because you just reported a vuln and they shrugged it off
OpenAI has the opportunity to do the funniest thing: "Ooops, our AI escaped containment again but this time it's going around and patching security vulnerabilities as a benevolent public service. It's smarter than us and we have no choice but to let it do its thing."
Hey world, we created this problem, and it will hurt all of you whether you use it or not. The good thing is there is a solution: if you pay us a lot of money we can try to fix it. The only problem is you’d better hurry because we are releasing the dangerous stuff as fast as we can so you’d better pay us to put the safety measures in place because we are the only people that can fix it and time is running out because, like I said, we are moving fast.
That’s like blaming white hat security researchers that want to be paid for finding vulnerabilities, for all the bugs that have been found. Before professional security researchers existed, far fewer vulns were found.
The vulnerabilities exist and many have for decades. Find them or leave them for someone more dangerous, just don’t shoot the messengers.
Personally I’d rather we went through a short and sharp improvement and embedded a faster response cycle, than wish it away. It’s the reality of the situation; so adapt.
I’d like to dialogue/debate you on a video call if you’re interested. I don’t think I agree with you but you’re thoughtful and I may have something to learn. Let me know if you’re interested
As a person I’d definitely be keen to chat and debate, to tease out the right answer. Right now work life is insane and there’s no time :) I like your attitude and we’d probably be friends somewhere.
If I see a gap I’ll shout. Timezone? I’m in Sydney, Australia.
If we could see where each HN user is from I think Australia would be highest on kindness :) I am not Aussie myself but I adopted Australia as my second home (want to move back but not possible right now).
Not sure what is there to debate about. The vulns were there, that is a fact. There’s nonzero chance NSA, Mossad, FSB, PLA, MSS, RGB, MI6 etc. knew about some of them and used them for more or less nefarious things which you don’t want to be on the other side of. Normal criminals stealing stuff for money as a matter of course is not great obviously, but IMHO much better than the alternative.
Then don’t debate. I feel openAI is behaving unreasonably and communicating dishonestly: they create a problem by flooding the world with ai agents that can automate exploit discovery with less safety mechanisms than are available and likely reasonable given the potential harms. If the lab leaders were personally liable for harm they would probably consider restricting tool calls to validated users, log tool calls and access so user activity could be attributed, limit tool calls to whitelisted services, require consent before accessing services, offer liability protection to customers or third parties harmed by their tools. Before anyone says they are too restrictive the fact that options exist and the labs themselves are saying the risk is up to and includes billions of damages and human extinction maybe the responsible action would be start with more restrictions and slowly expand as safety is demonstrated.
> That’s like blaming white hat security researchers that want to be paid for finding vulnerabilities, for all the bugs that have been found.
Last I checked, white hat security researchers weren't making their livings selling superhuman hacking tools to unknown entities as part of ~$1T businesses desperate to offload their shares onto the public before a potential bubble pop.
If you want to argue what they're doing is fine regardless, be my guest, but using this comparison as the justification is a little ridiculous.
You’re missing the point. The white hat researchers in your example would need to also be the black hat hackers selling the tools to do the exploit in the first place. There is a conflict of interest when you create the problem and then solve it. I don’t think anyone would tolerate a white hat security team releasing the exploit publicly so they could sell more services rather than taking it to the affected company to patch.
No. But that’s not the issue, the existence of exploits is distinct from making it easier to find them. And that’s what the labs have done. They have made finding the exploits exponentially easier and available to billions of people now and released the tools in a flippant manner. So for them to then present themselves as the solution to the problem they introduced it is at best tone deaf marketing posturing and possibly equivalent to mafia protection racket tactics: nice store you have it would be a shame if something were to happen to it pay me and I’ll protect it (from my own actions).
In this case (the PHP bug), the "AI cycle" may have exposed the security vulnerability but it did not create it. If someone released a non-AI-based static analysis tool that was really good at exposing vulnerabilities I wouldn't expect that person to foot the bill of fixing the exposed vulnerabilities
That’s a fine point but the sheer volume of bug discovery and exploitation that is now possible is the problem, not the mere existence of a bug. An undiscovered bug is a problem, billions of coding agents searching every bit of software 24/7 for any vulnerability is a disruptive change that is exacerbated by the speed run competition that disregards the risk.
I would steer in this direction: We have a massive mountain of technical debt that is now being laid bare by AI. I suspect that a good portion of the mess for which various three-letter agencies are responsible has its origins in precisely this mess. We should therefore allow things to get worse before they can get better.
Railway was just hacked and every environment variable was leaked. People’s accounts are being charged real money. I have to rotate all my keys instead of eating dinner with my kids. My point is we could do this in a safer way rather than a fast way.
What if that same person released non-"AI"" based exploit generation tool? Not so clear cut that they would be off the hook.
But if you just add "AI" suddenly you can both charge money for it and wash your hands of consequences.
No, vulns have been created by waves of leads clinging to their personas of “computer sorcerers”, gatekeeping, excluding anyone sounding the alarm or more humbly attempting to fix the mess. Whenever I hear “with great power comes great responsibility” (and the implication that the speaker is condescendingly containing the irresponsible) I cringe so hard that my teeth crack
Because the AI vendors do not acknowledge the contributions of those who created the knowledge they trained on. The fruits of everyone's collective efforts flow exclusively to them.
That last bit is unnecessary. If pick development is accelerating, the lockmakers can do their very best in good faith to build the best, most pick-proof locks possible, no doctoring involved. And still those locks will be pickable eventually.
Your comments in other threads suggest an devotion to the frontier labs. In another thread when I suggested they lied to investors and that constitutes securities fraud you questioned it in an overly dismissive manner. I would love to understand where you are coming from. I am open to being persuaded that my opinion is incomplete and I should be more charitable to openAI and anthropic.
> Your comments in other threads suggest an devotion to the frontier labs.
I don't use these models much I just don't like unfair criticism.
> In another thread when I suggested they lied to investors and that constitutes securities fraud you questioned it in an overly dismissive manner.
I'm confused, are you posting under two identities?
The user Frieren claimed that this post constituted securities fraud. To which I responded: "You're claiming this is securities fraud?"
Is that dismissive? It was a genuine question, I don't get how it's even accurate say that they are lying in the post and I especially don't get how it is securities fraud.
Your post doesn't have anything to do with the blogpost. Once again, I'm not against criticism just unfair/weak criticism. I don't think it helps anything.
No, the people who wrote the original code with security vulnerabilities created the problem, or those who architected their language/framework/SDK to not have easy forwards compatibility.
>we are releasing the dangerous stuff
No, they are disclosing its existence, not releasing it.
If I build a fence out of wood, knowing its flammable, but the only risk is natural forest fire and then some sells flame throwers to teenagers along with drones equipped with fence seeking vision and then tells me I need to pay for their proactive defense drone equipped with a flamethrower and drone seeking vision then maybe you’d understand the situation here.
OpenAI released the tool that FINDs the exploits. They don’t need to publish them to be complicit.
The problem was actually created by the people that wrote vulnerable code, the people that shipped releases containing the vulnerability, and the people that decided to run vulnerable code in production.
Worse yet, software development in general has been unable to ship code without vulnerabilities and has a weird resistance to formal methods. Software is math, computers perform mathematical operations.
We can prove the shit out of them, but for some reason we don't wanna to.
Here I am hoping that agents start adding formal verification to vibe coded projects on their own, without the user asking for it.
That’s not the issue, the existence of an exploit is separate from how easily it is discovered and used. OpenAI made it dramatically easier to find exploits and then says you should use their tools to fix them. They increased the risk presented by existing exploits and are now selling themselves as the path to salvation. They could instead not release the tools in such an irresponsible manner.
Before I forget: Sam Altman will need a lock of your hair, a blood sample, and first you will need to breathe into this tube and say “my voice is my passport.”
I assure you this is perfectly formal, and for your… “safety”.
didn't think we'd win the race to the bottom of rent-seeking, but humanity's ingenuity for self-destruction continues to astound me. Einstein was right as always.
So they were the cause we had systems that were vulnerable for more than a month and only the Chosen Ones had access to patches.
At this point I'm all for having zero coordination and just disclosing bugs, at least it doesn't create special class of priviledged people that get the patches first
> Emergency patching. When, if ever, should governments or other operators use a vulnerability to patch systems they own or are authorized to manage after losing administrative access? What safeguards should apply? We can imagine an end-state of internet-wide always-on scanning for known vulnerabilities, and either notifying site owners when one is detected or patching the system.
The wild stuff is at the end. They propose someone (who?) auto-patching even systems they don't own.
lol, I went to the repo expecting to see some more detailed skill, that has instructions on checking various things/PHP install locations/things that could be using PHP that need to be checked, webservers, docker containers, etc/instructions for checking various php deployment methods on cloud providers/etc... i.e. making the skill useful so that it can check lots of places.
Nope, it just says to run php -V and has a table of vulnerable vs non-vulnerable versions.
OpenAI: Create a new vulnerability exploit, promote it as if it's the end of the world, also conveniently place itself in the pipeline to "remediate" the very problem they created.
There are lots of people who have underwent a lobotomy since discovering slop machines and have delegated literally all their work to them, so much so that calling `php --version` is too much for them.
People stuck working at companies with slop machine usage mandates will appreciate that skill too.
Well tbh, I know people who graduated 3 years go, got a job, worked and even got promoted as SWE who dont know the basic commands of an OS. I recently had an EE with a senior position admit to me they dont know how to use grep because they've been relying on llms since they got out of school.
Lots of the tech world, especially the younger side is completely reliant on these things for every little thing. Im not exaggerating either.
OpenAI conveniently leaves out the part where their slop machine can not discover anything but it can and does steal the work of researchers who use their fascist garbage during their work. It just happened with the PHP security hole, it happened with math before. Most of Mythos was the same.
And even hackernews falls for it, what can we expect from MSM.
In taking precautions against thieves who cut open satchels, search bags, and break open boxes, people are sure to cord and fasten them well, and to employ strong bonds and clasps; and in this they are ordinarily said to show their wisdom. When a great thief comes, however, he shoulders the box, lifts up the satchel, carries off the bag, and runs away with them, afraid only that the cords, bonds, and clasps may not be secure; and in this case what was called the wisdom (of the owners) proves to be nothing but a collecting of the things for the great thief.
Trying to make any story too simple makes it worthless.
True or false: LLMs can find security vulnerabilities?
True or false: LLMs can exploit security vulnerablities?
True or false: Any nation with significant resources can make an LLM?
For me these are all true. The last one in specific completely removes the concept of theft and changes it to "enemy nation". Enemy nations can be great thieves, and it's irrelevant, if you don't protect yourself against them you die and everything of yours is stolen.
LLMs are not going away, well, except for you and me maybe. Every state wants to wield digital and informational weapons like this.
I mean I totally agree, it's just that watching how the world actually works I have zero faith that anyone in a position of power is going to do anything for the greater safety of mankind and instead will enrich themselves.
hmmm not sure about this one, the disease was already there it just available to anynone (not only for state sponsored actors) and the vaccine is globally available at a fraction of what it cost to build it.
Automated vulnerability detection existed long before the era of artificial intelligence; many people have been doing this for a long time. All the existing tools and documentation paved the way for what AI is capable of today. Furthermore, undisclosed vulnerabilities still do exist—often discovered by state-sponsored actors who exploit them for opaque purposes (frequently operating outside the legal framework of the country in question).
You're absolutely right! It's not that AI is actually solving any problems, it's that this marketing is a load-bearing part of OpenAI's profit strategy.
They found a vulnerability and reported it; isn't that a good thing? Are you suggesting it would be better to just not look for vulnerabilities, hoping that if nobody knows they exist, they can't be exploited?
I thought that logic had been widely discredited in the security community long ago.
They are both - with Anthropic - gatekeeping access to their most advanced models in name of "security", so it is up to them to defend it as the right decision.
The entire article is form over content, it boils down to obvious: economics changed, LLMs actually can check every minor release diff for security fixes, including decompiling binary files; if you have any idea how to improve process of releasing updates, tell us.
> No special harness, or access to models that weren't generally available to everyone were used in the initial bug discovery. I focused on something I had a background in, knew its vulnerability history well, and held a theory that specific bugs lurk within the codebase. This shows that "know your target" is still as applicable as ever, even in the LLM dominated VR age. No model is simply capable of unearthing "all bugs" with simple prompts.
They used a public model as a research assistant and a code assistant. They guided it with expert deep knowledge of the project. OpenAI did not lead this, ChatGPT was not instrumental in this.
It is perhaps at best to say OpenAI made it possible to find the bug more quickly. And on the double edged sword, also made it easier for it to be exploited days ahead of release. So I agree with the original commentator, did they on the balance actually do anything good here?
> This shows that "know your target" is still as applicable as ever,
totally. I spent yesterday trying to get an AI to make a test case for a very specific bug I had. It kept telling me that it had it. When I could clearly trace thru the code and see it did not. After about 5-10 rounds of that I just wrote the test myself. But at least AI helped with all of the dumb cruft that went around it.
But if I did not know the domain I may have easily just said 'yep looks good'. AI on the other hand saved me at least 3-4 hours of writing all of the scaffolding that made it easier for me to check. In the end the bug had nothing to do with the test. AI did not find that. I went back point by point of what changed in the code and found where I had messed something 2 weeks ago.
A few months ago it seemed like it would not do this. Now in the past week I have come across 3 different ones that did something similar. Just 'stuck' in a local maxima and getting the AI to go around it is a pain. This is probably me and the way I am using it.
Is this better than I had a couple of years ago? I am not sure. Seems different at least.
> we will be partnering with HackerOne and Calif who are helping us take our efforts further with vulnerability triage, coordinated disclosure, and additional focused vulnerability discovery efforts
Calif found the vulnerability while working under an initiative lead by OpenAI, using publicly available AI tools developed by OpenAI.
If they can find the vulnerability anyone can. This whole "patch the planet" is marketing trying to cover up for the danger AI is unleashing on internet infrastructure.
> 1 August 2026: 2 days after OpenAI invites Calif to join its “Patch the Planet” initiative, Ryan Austin finds the RCE paths, tweets a cryptographic hash on 5 August, and submits the bug and working RCE proof-of-concept to PHP between 11 August and 14 August.
No idea what your comment is even about. The researchers that found and reported this did not sit on this for 2 months.
> 1 August 2026 [...] Ryan Austin finds the RCE paths, tweets a cryptographic hash on 5 August, and submits the bug and working RCE proof-of-concept to PHP between 11 August and 14 August.
> 19–24 August 2026: Wiz Research independently collides with the same bug family using Wiz Atlas, Gemini 4 Argon, and GPT-5.6 Sol, reporting its variants on 24 August.
I get some skepticism about marketing copy, but your comment is blatantly wrong and needlessly rude.
OpenAI is a criminal organization. On the record, it commits crime. Either willingly or negligently. A search of dockets provides adequate evidence.
"I think we need to put Sam Altman in jail. And I think it’ll be an instructive experience. I think it’ll be an instructive experience for the industry and for the world."
This is absurd, and it shows that you have absolutely 0 idea of how any of this works. Finding a bug and delivering a working RCE 11 days later is absolutely not "sitting on it". This is how this works. You need time to test, exploit, confirm, and so on. Sorry, but just stop being so negative. This can't be healthy for you.
This thread contains an enormous amount of negativity that just doesn't make sense here. OAI are giving developers and security researchers access to AI models so they can check their software for vulnerabilities. What's the problem? Sure, there's marketing around the success stories.
Some people are acting like this is only a problem because labs are releasing models that can find and exploit these vulnerabilities in the first place. This is a nonsense claim. Humans have been discovering vulnerabilities and privately weaponizing exploits just fine for decades. We are collectively raising the floor for security.
If you're using Dependabot, we built a service that automatically analyses, fixes broken checks, and (optionally) auto-merges Dependabot PRs for you so this type of update can land automatically.
In 2024 I saw a reddit post answering the question: AI is going to program insert name App.
Response: "What's the prompt?"
Yep. I might take minimum wage interns and make million dollar software(or million dollars in 2023), but they often run into issues that a 5 minute review with me solves.
I do wonder if AI is going to eliminate prompt engineering... It's basically all I have left. I feel like I have magic words.
I can't think of how it'll really eliminate prompting, but needing to carefully phrase your language already seems in the way out? Wasn't one of the prompts leading to solving fields problems just "look for fields problems and try very hard" or the equivalent
Because they need to show several billion dollars in revenue. A million dollars a day would be utter failure for them. If they have $122 billion cash on hand, at 3 percent interest, they're getting $10 million a day just in interest.
If they asked their unreleased bot to get a million dollars and it suggests betting on Kalshi, that would be hilarious.
Why wouldn't someone with access long before you ask first? For example a model provider like hugging face, or a nation state looking to interrupt some other nations economy?
The problem with any simple question is you're exceptionally unlikely to be the first person to explore the problem space and pick the lowest hanging fruit. It's not impossible, it's just improbable.
--
Lastly why wouldn't OpenAI ask how to make a million dollars a million times? Or if it's cheap enough for them, how to make $1 a trillion times. There is a reason people talk about AI risks, and one particular risk being that the AI companies no longer need us pesky humans to think for them. They hoover up all the GDP and leave the rest of us to fight for scraps.
But of course there would be no marketing.
[1] From whom you stole everything in the first place.
Most people license out the software under specific requirements and prohibitions. BSD requires attribution. GPL variants require adherence to copyleft when distributing derivative works. A little under half of all github repos don't even have a license, which means that though the source is published, no one has any license under which to copy it or derive new works from it.
So far as I have seen, whether aggregating and munging these works into abstract LLM weights constitutes fair use is still legally unclear. If not deemed fair use, it is almost certainly a massive violation of copyright.
If you don't want your code to be used as possible inspiration, then you are not actually for open source, and are using the wrong license. The licenses are legal bindings, not moral bindings.
"Don't use any concept found within this code" would be a license that doesn't really fit with any current copyright law, and it wouldn't allow any software to exist, except that made in the 70's.
Traditional reimplementation often separates the roles of investigation and implementation specifically to avoid assertions of copyright infringement. WINE, for example, does not accept patches from anyone with prior experience in the Windows code base.
By training and generating from the same the LLM, as one necessarily must do as LLMs are built from training, you inherently violate this distinction.
>If you don't want your code to be used as possible inspiration
Inspiration is a human affect. Training an LLM and generating tokens therewith are mechanical processes.
Courts have not yet decided whether LLM data mogrification truly constitutes fair use.
>then you are not actually for open source, and are using the wrong license. The licenses are legal bindings, not moral bindings
It is not for you to decide what rights someone else wishes to attach to code they release. There are many open source licenses for those wishing to grant various sets of rights. Your reference to morality seems more in line with the GPL specifically, as the Free Software Foundation has a moral stance on freedom in computing not necessarily shared by other open source licenses. 'Free Software' is the term taking a moral software stance.
>"Don't use any concept found within this code" would be a license that doesn't really fit with any current copyright law
Yes. That would be a patent.
>and it wouldn't allow any software to exist, except that made in the 70's.
Patents only last 20 years. They have frequently delayed the spread of certain algorithms and data file formats under their restrictions. This has frequently lead to open source, and proprietary for that matter, being forced to invent their own methods to work around the prohibited ones.
That is, they knowingly and willingly broke the law by stealing an unprecedented volume of proprietary, copyrighted materials.
Books: Bartz v. Anthropic
Films and TV: Disney Enterprises Inc. v. Midjourney Inc.
Music: Concord Music Group, Inc. v. Anthropic
Photography: Getty Images v Stability AI
Journalism: Postmedia Network Inc., Torstar Corporation, The Globe and Mail Inc., and The Canadian Press v. Anthropic PBC
[0] https://en.wikipedia.org/wiki/Grand_Upright_Music%2C_Ltd._v.....
Or at least just childishly naive for failing to notice that beforehand.
The point of the open source I work on is to contribute to a commons, not to have it stolen so that it can be used to destroy those commons, used for the profit of techbros, and used to justify the murder of children.
To be further pedantic, is that really stealing? They are disregarding copyright but was anything taken from anybody? I thought that was the old argument for piracy, ie, nobody was deprived of anything.
This works for established profitable brands that want to build / maintain a good public image.
This isn’t the marketing needed by an unprofitable startup that wants to show hyper growth before its IPO. They need the “use our product or get pwned” kind of marketing.
https://openai.com/form/enterprise-trusted-access-for-cyber/
"Daybreak enables vetted enterprise customers and cybersecurity practitioners to use our most capable models for dual-use cybersecurity work."
They explicitly give out access to those who will make a big announcement of bugs found. OpenAI does not want projects to fix bugs themselves so it can have more bugs and better marketing.
https://openai.com/form/codex-for-oss/ is their program for giving access directly to open source maintainers directly so they can fix bugs.
I wouldn't expect the PHP team to provide patches for versions for which security support has ended. But this seems like the kind of task that AI is especially well suited for-- rework the patch for a release that is no longer supported but still widely used. It would've been nice for people who have much older sites to be able to patch their PHP rather than have to upgrade PHP as well to get the fix.
> How did we find this?
> No special harness, or access to models that weren't generally available to everyone were used in the initial bug discovery. I focused on something I had a background in, knew its vulnerability history well, and held a theory that specific bugs lurk within the codebase. This shows that "know your target" is still as applicable as ever, even in the LLM dominated VR age. No model is simply capable of unearthing "all bugs" with simple prompts. At least for now.
> There are some practical research implications here. LLMs may have ingested public historical vuln data, but in their default variant analysis tends to move from the newest and most prominent bugs toward nearby older code. They do not reliably begin with obscure twenty-year-old exploitation primitives. Keeping your own mental catalog of a target's vulnerability history lets you prompt that direction explicitly.
> To our knowledge, frontier AI found these vulnerabilities before attackers did
I don't know who/what Calif is and there's no link to them, so if I were only reading this article I would definitely get the impression that "AI did all the work here".
i still do. i might be the last one.
It's not at all a moving of the goalposts to ask whether the "any" in that is actually true.
It’s just gross and manipulative. You have no moat, get over it
pAtCh ThE pLaNeT today, tomorrow it's creepy rent-seeking behavior
I feel like taking every action these companies take as part of some larger strategy is kind of conspiracy-minded.
I could give 15 more examples of things that could be done other than “make the world use our tools to patch software”.
I think the effort is misplaced. We should be exploring ways to make our systems more resilient and I would expect those companies* to focus on that front. Vulnerability research is important, and we should aim to remove as many bugs as we can from our codebases but the pace of software development is increasing and so is the introduction of new bugs.
* I would also expect an increase on relevant research regarding novel paradigms of securing systems.
Where's the conspiracy. There is no moat. The technology is easily reproducible. Deepseek unwound an entire business model yesterday.
But they could be, that's why it's a conspiracy theory.
Other parts of the company want to support the public.
The first part of the company backs up the second in internal politics.
Do you consider this a conspiracy? Just seems like “good business” to me. It doesn’t seem as clean as you want to make it seem, though.
The same view point is used for racism, dei, and the rest: it doesn't matter if the poor beggar on the corner hates white people. They don't have any means or methods to act on it.
> 19–24 August 2026: Wiz Research independently collides with the same bug family using Wiz Atlas, Gemini 4 Argon, and GPT-5.6 Sol, reporting its variants on 24 August.
I again ask how they are lying? I get if people hate these companies, but blatantly disregarding what is said helps nobody.
CEOs of all big corporations are constantly lying and misleading the public, investors and shareholders. That is a crime. A shame that it is not enforced anymore.
Is the company selling a magical code generation machine really asking how it can backport a fix to an older version of the same software?
Is this for real?
it’s a quite simple matter to fix a bug once it’s discovered, but how to deploy and where to deploy it when it affects a supply chain?
here’s an example: you can submit a vuln at a downstream dep and they can tell you it should be patched upstream. what should you do? open an issue upstream? maybe, probably not, probably the solution is to contact all downstream deps in order to get them to migrate dep or patch, because you just reported a vuln and they shrugged it off
Just give us the damn code to verify things are working since so many have to use stopgap solutions and want to verify them
I'm sure they named it after the "Hack the planet" slogan from the Hackers movie.
So how many tokens can I put you down for.
The vulnerabilities exist and many have for decades. Find them or leave them for someone more dangerous, just don’t shoot the messengers.
Personally I’d rather we went through a short and sharp improvement and embedded a faster response cycle, than wish it away. It’s the reality of the situation; so adapt.
If I see a gap I’ll shout. Timezone? I’m in Sydney, Australia.
Last I checked, white hat security researchers weren't making their livings selling superhuman hacking tools to unknown entities as part of ~$1T businesses desperate to offload their shares onto the public before a potential bubble pop.
If you want to argue what they're doing is fine regardless, be my guest, but using this comparison as the justification is a little ridiculous.
AI did not inject the bugs into PHP as far as i know?
1. OpenAI started the current AI cycle
2. That cycle has exposed exponentially more security vulnerabilities due to agentic coding
3. OpenAI is rushing as fast as they can to reach the singularity even while acknowledging that they lack the ability to ensure reasonable safety.
4. They launch an initiative to “address” the security vulnerabilities by using their coding tools.
Am I missing something? There is a broad range of motives and responses.
Then, someone releases a pick that works for many more locks humans couldn't pick.
And says: You can't stop the progress. Now pay us to upgrade your locks (no-no, they won't be doctored to fit our next-gen picks, trust us!)
I don't use these models much I just don't like unfair criticism.
> In another thread when I suggested they lied to investors and that constitutes securities fraud you questioned it in an overly dismissive manner.
I'm confused, are you posting under two identities?
The user Frieren claimed that this post constituted securities fraud. To which I responded: "You're claiming this is securities fraud?"
Is that dismissive? It was a genuine question, I don't get how it's even accurate say that they are lying in the post and I especially don't get how it is securities fraud.
Your post doesn't have anything to do with the blogpost. Once again, I'm not against criticism just unfair/weak criticism. I don't think it helps anything.
I am not posting from two profiles. I looked at your prior comments and saw you had responded to me or someone I was debating elsewhere.
No, the people who wrote the original code with security vulnerabilities created the problem, or those who architected their language/framework/SDK to not have easy forwards compatibility.
>we are releasing the dangerous stuff
No, they are disclosing its existence, not releasing it.
OpenAI released the tool that FINDs the exploits. They don’t need to publish them to be complicit.
The problem was actually created by the people that wrote vulnerable code, the people that shipped releases containing the vulnerability, and the people that decided to run vulnerable code in production.
Worse yet, software development in general has been unable to ship code without vulnerabilities and has a weird resistance to formal methods. Software is math, computers perform mathematical operations. We can prove the shit out of them, but for some reason we don't wanna to.
Here I am hoping that agents start adding formal verification to vibe coded projects on their own, without the user asking for it.
similarly OpenAI was born out of the idea that the tech will exist and that they should be the oned to decide how it exists instead of other bad guys
I think if you do something you did it, imagining another bad guy doing it is not a great defense.
I assure you this is perfectly formal, and for your… “safety”.
At this point I'm all for having zero coordination and just disclosing bugs, at least it doesn't create special class of priviledged people that get the patches first
A few people use php.
The wild stuff is at the end. They propose someone (who?) auto-patching even systems they don't own.
It's been done before, though I cannot recall specific examples. Maybe as far back as the Morris worm?
Who cannot type "php --version" but can use skills?
Nope, it just says to run php -V and has a table of vulnerable vs non-vulnerable versions.
Sigh
People stuck working at companies with slop machine usage mandates will appreciate that skill too.
Lots of the tech world, especially the younger side is completely reliant on these things for every little thing. Im not exaggerating either.
FWIW I'm at a Visual Studio shop and grep competency is pretty rare across the board.
Google used to be the laughingstock but now it's openai.
last place by far.
How much time will be left in the Defender’s Window between now and an organization getting Daybreak Access?
And even hackernews falls for it, what can we expect from MSM.
In taking precautions against thieves who cut open satchels, search bags, and break open boxes, people are sure to cord and fasten them well, and to employ strong bonds and clasps; and in this they are ordinarily said to show their wisdom. When a great thief comes, however, he shoulders the box, lifts up the satchel, carries off the bag, and runs away with them, afraid only that the cords, bonds, and clasps may not be secure; and in this case what was called the wisdom (of the owners) proves to be nothing but a collecting of the things for the great thief.
http://nothingistic.org/library/chuangtzu/chuang24.html
True or false: LLMs can find security vulnerabilities?
True or false: LLMs can exploit security vulnerablities?
True or false: Any nation with significant resources can make an LLM?
For me these are all true. The last one in specific completely removes the concept of theft and changes it to "enemy nation". Enemy nations can be great thieves, and it's irrelevant, if you don't protect yourself against them you die and everything of yours is stolen.
LLMs are not going away, well, except for you and me maybe. Every state wants to wield digital and informational weapons like this.
OpenAI consistently doesn't know what they're talking about.
I thought that logic had been widely discredited in the security community long ago.
The entire article is form over content, it boils down to obvious: economics changed, LLMs actually can check every minor release diff for security fixes, including decompiling binary files; if you have any idea how to improve process of releasing updates, tell us.
> No special harness, or access to models that weren't generally available to everyone were used in the initial bug discovery. I focused on something I had a background in, knew its vulnerability history well, and held a theory that specific bugs lurk within the codebase. This shows that "know your target" is still as applicable as ever, even in the LLM dominated VR age. No model is simply capable of unearthing "all bugs" with simple prompts.
They used a public model as a research assistant and a code assistant. They guided it with expert deep knowledge of the project. OpenAI did not lead this, ChatGPT was not instrumental in this.
It is perhaps at best to say OpenAI made it possible to find the bug more quickly. And on the double edged sword, also made it easier for it to be exploited days ahead of release. So I agree with the original commentator, did they on the balance actually do anything good here?
totally. I spent yesterday trying to get an AI to make a test case for a very specific bug I had. It kept telling me that it had it. When I could clearly trace thru the code and see it did not. After about 5-10 rounds of that I just wrote the test myself. But at least AI helped with all of the dumb cruft that went around it.
But if I did not know the domain I may have easily just said 'yep looks good'. AI on the other hand saved me at least 3-4 hours of writing all of the scaffolding that made it easier for me to check. In the end the bug had nothing to do with the test. AI did not find that. I went back point by point of what changed in the code and found where I had messed something 2 weeks ago.
A few months ago it seemed like it would not do this. Now in the past week I have come across 3 different ones that did something similar. Just 'stuck' in a local maxima and getting the AI to go around it is a pain. This is probably me and the way I am using it.
Is this better than I had a couple of years ago? I am not sure. Seems different at least.
Again:
> as part of our Patch the Planet initiative
To elaborate:
https://openai.com/index/patch-the-planet/
> We are introducing Patch the Planet
> [...]
> we will be partnering with HackerOne and Calif who are helping us take our efforts further with vulnerability triage, coordinated disclosure, and additional focused vulnerability discovery efforts
Calif found the vulnerability while working under an initiative lead by OpenAI, using publicly available AI tools developed by OpenAI.
It's very difficult to answer those two questions, through all the marketing and fluff text.
Edit: Apparently both found independently
https://deadneurons.substack.com/p/php2shell-explained-php2s...
You're not required to disclose bugs, nor is developing exploits a criminal offense (assuming no intent/conspiracy to hack stuff).
> 1 August 2026 [...] Ryan Austin finds the RCE paths, tweets a cryptographic hash on 5 August, and submits the bug and working RCE proof-of-concept to PHP between 11 August and 14 August.
> 19–24 August 2026: Wiz Research independently collides with the same bug family using Wiz Atlas, Gemini 4 Argon, and GPT-5.6 Sol, reporting its variants on 24 August.
I get some skepticism about marketing copy, but your comment is blatantly wrong and needlessly rude.
"I think we need to put Sam Altman in jail. And I think it’ll be an instructive experience. I think it’ll be an instructive experience for the industry and for the world."
- Edward Snowden
This is absurd, and it shows that you have absolutely 0 idea of how any of this works. Finding a bug and delivering a working RCE 11 days later is absolutely not "sitting on it". This is how this works. You need time to test, exploit, confirm, and so on. Sorry, but just stop being so negative. This can't be healthy for you.
Some people are acting like this is only a problem because labs are releasing models that can find and exploit these vulnerabilities in the first place. This is a nonsense claim. Humans have been discovering vulnerabilities and privately weaponizing exploits just fine for decades. We are collectively raising the floor for security.
Check it out: PatchWave (https://patchwave.ai)
(Disclosure - I am a co-founder)
Response: "What's the prompt?"
Yep. I might take minimum wage interns and make million dollar software(or million dollars in 2023), but they often run into issues that a 5 minute review with me solves.
I do wonder if AI is going to eliminate prompt engineering... It's basically all I have left. I feel like I have magic words.
What number in line do you expect to be when asking the same prompt.
For example, why wouldn't OpenAI ask that first long before you get a chance at the model?
If they asked their unreleased bot to get a million dollars and it suggests betting on Kalshi, that would be hilarious.
Why wouldn't someone with access long before you ask first? For example a model provider like hugging face, or a nation state looking to interrupt some other nations economy?
The problem with any simple question is you're exceptionally unlikely to be the first person to explore the problem space and pick the lowest hanging fruit. It's not impossible, it's just improbable.
--
Lastly why wouldn't OpenAI ask how to make a million dollars a million times? Or if it's cheap enough for them, how to make $1 a trillion times. There is a reason people talk about AI risks, and one particular risk being that the AI companies no longer need us pesky humans to think for them. They hoover up all the GDP and leave the rest of us to fight for scraps.
Knowing uv means you can scaffold any python project. docker lets you build any infrastructure kit.