I'd argue that the resources (i.e. time, money and effort) needed to create the tooling that would have prevented all of these issues would have been orders of magnitude less than the resources that have gone into "AI".
It’s a bad argument. AI isn’t single use. Current iteration of AI is already there and serving it is cheap. So you get great tool for detecting these issues in the future at low cost. Without AI the cost wouldn’t really decrease with time. And also it costs me 200$ now to find these bugs. I don’t think I could hire software engineer for that much.
This said memory unsafe code just likely has more bugs because of all the footguns it has on top of all the other security work you need to do with it.
The orders of magnitude fewer security vulns for OpenBSD are not due to less testing. Everybody has always known OpenBSD is more secure, and it's not from using a different language. The reason is the code. Linux was not designed to be secure. OpenBSD was. It's that simple.
As an OpenBSD lover and frequent user I will tell you it does not support but a significant fraction of the hardware Linux does, and has many features that are reduced in functionality or just gone compared to Linux.
Is it bulletproof? Definitely! But you need the genius of Albert Einstein as well as the patience of Mother Theresa to actually use it in production.
For most server setups OpenBSD's hardware support is adequate, and this is especially true for VMs. It's laptops, peripherals, and bleeding edge, high-performance or specialized devices where driver support is a PITA. But that's not something I've encountered running my own private servers. It would be nice to have something like ZFS, but mostly that's a peace of mind and convenience thing--would be nice to be able to stream backups somewhere rather than push around dumps and tarballs, though I have a script for that.
More than half of those lines are from drivers, drivers which wouldn't be relevant to a VM environment. Core kernel code for any particular architecture is less than 10 million, but some of that is filesystems or subsystems not in use or available to an unprivileged user on a typical server, so possibly maybe 5 million or less. And the article says they only targeted 2.8 million SLoC.
I think this is basically true, but it didn't "die" in the sense that we gave up on it; it just never worked. Systems software was much worse in the 1990s than it today; you could breathe on something with a string of AAAA's and get a 0x41414141 register core dump.
> ... it didn't "die" in the sense that we gave up on it ...
It died in the sense that the "people in charge" hired us for our expertise and then ignored our every suggestion and fought us tooth and nail every inch of the way when we suggested that security or reliability ever mattered in the slightest until after the disaster happened, and then straight-up scapegoated us for "not pushing back harder" against the threats of "go work somewhere else then, code-monkey!" until we gave up and did the "code-monkey" thing for our paycheck and no longer cared about quality or "a job well done".
It died because personal computers got too powerful so nobody bothered designing and building “true” multiuser systems (that bored college students were trying to break).
Back in the 90s, it's not as if any semi competent cs student couldn't slice your typical college server like swiss cheese. People thought null terminated strings were fine back then, and patching everything immediately was rare. So multi user was already dubious.
> Should we assume that all software is more compromised than the kernel?
Absolutely, the front door has always been unlocked. We were previously protected by the economics of exploit detection favoring the developers, but now we have entered a new world.
The issues arise almost everywhere we look. It's frustrating because the user experience in edge systems has barely changed in the last fifteen years. We could have rode the advances in semiconductor manufacturing to embed more layers of protection such as CHERI while slightly slowing down the rate of performance improvements. Instead we have only marginally upgraded security and largely added features which the users scarcely notice.
I am hopeful about the long term prospects of formal verification, but having humans reviewing the proof statements and confirming they match our desired properties will take many years. Meanwhile, AI assisted hacking is an urgent issue. So we will have to delegate implementation, proof specification, and proof development to the AIs in the short term. Maybe that goes well, but who knows. Interesting times.
I recall reading that every IBM mainframe OS release had 1000 bugs in it. (This was in the 1970s/80s.) Sorry I can't supply a source, after 4-5 decades. It was probably an article in Datamation magazine.
Now, I don't know how many of those bugs were privilege escalations, or could be leveraged to be. I'm not sure that people tried to exploit in the same way back then (or, if they did, if they said anything publicly about the results). Still, even if the majority of those bugs were not exploitable, OSes were much smaller then.
So that's my conclusion of how we got here: We've always been here.
Mainframes went right to virtual machines, so possibly bypassed much of the multiuser faux-security which has always plagued unix. Of course, now they run Linux partitions and get all the same bugs.
For all of my time being aware of this industry, it has generally been understood that privilege escalation bugs are common. This has always been a thing. A lot of people say, commonly and for many years, that if your security depends on there not being privilege escalation bugs, you don't have security.
The kernel has the problem that basically every bug in it is also a security vulnerability. All software has bugs. Had Linux been written in Rust, and I strongly recommend a rewrite, it would still have bugs, and thus would still have vulnerabilities. But they'd be way less.
fwiw at work we've been looking at what the options are more memory-safe kernels out there. Not many options. If there were a "safe linux" maintained by like canonical or redhat or whatever we'd use it. For datacenter workloads like sandboxes and virtualization I think it'd do well.
Won't AI's ability to discover these bugs reduce the need to rewrite things in rust? If the major benefit to rewriting them is memory safety, I mean...
That's assuming the current models are able to find all the bugs, and that we won't have a better model out every few months that will be able to find even more obscure bugs.
We would also have to burn a trillion tokens a day verifying all the C code the hard way forever going forward. It would be cheaper to rewrite the kernel in Rust, subsystem and driver by driver at a time.
That's a fair question, and actually one I ask myself too.
AI doesn't give a strong guarantee that things work though, while Rust gives a much stronger (not absolute) guarantee.
But maybe the way of the future is to auto-generate proofs that the code is memory safe and then just leave things where they are, instead of doing rewrites that require maintainer buy ins. The advantage of C plus proofs would be that the AI generated proofs could live separately so it wouldn't touch workflows of the maintainers.
The only required output they have to process is evidence of vulnerabilities in changes proposed to them. They could outsource even that to contributors.
Humans manually typing out the bytes of source code has been an outdated method, since around a year. It simply doesn't scale as much as those same humans directing agentic teams. So yes, it would be about prompts.
The hard part isn't the technology for the language transition, that one is present today. pgrust shows the way. It's ensuring maintainers are cooperative, i.e. a social problem. Eventually you diverge too much if you do a fork, so it's better done in cooperation with upstream.
Just check how long it took LLVM or GCC to switch off svn for example, and a language switch is much more involved than a VCS switch.
You mean the pgrust which was recently caught for cheating benchmarks and incapable of even performing basic queries outside the benchmark set? Would love to see pgrust debunked by Jepsen like this other AI-generated, written in Rust, "formally verified" scam.
I don't consider a cheater downplaying their cheating to resolve the fact that they cheated, no. Every cheater has a defense when caught, and theirs was not convincing. At this point I think the only thing that could redeem their brand would be paying for a Jepsen analysis to verify pgrust is as good as they claim it is; I have zero reason to trust the words from the horse's mouth.
1 bug per 10 lines -- not kilolines -- is much closer to the actual "bug rate".
You may be thinking "1 serious security vulnerability bug", but in my experience that too is much closer to 1 per 100 lines to 1 per 1,000 lines, depending on the type of application.
>Is the kernel simply too complex to reason about? Is it an underlying flaw in the security model or language implementation?
There is nothing magical about kernel. Look at the code anywhere and it has bugs, lots of them.
Security bugs are just bugs. More often than not they are silently there and don't prevent you from using the software and you can't react and fix them because of that.
Because the large corporations that fund Linux development and use Linux value speed over security.
Speed requires a lot of code in the kernel, where that code can run fast.
There are kernels that do not have a thousand privilege escalation exploits. These kernels are smaller and much easier to check for bugs. But those kernels are significantly slower.
For many use cases, slower is not a problem. But for the large corporations who fund a lot of Linux development and use Linux in large data centers, every 1% slower can literally be millions of dollars worth of additional servers that are required.
OpenBSD shows that you can have a Unix security model without a thousand privilege escalation exploits. But OpenBSD is like 10% slower. So corporations won't use it because it would cost millions of dollars more per year and some of their software requires access that OpenBSD does not grant precisely because that access leads to security exploits (i.e. io_uring, eBPF etc. etc.)
You could also use like seL4, but that's slow too
So that's why. It's a business tradeoff between cost of servers vs cost of getting hacked, and cost of servers wins.
It also used to just be fine patching the bugs as they were found. Its only recently where that model has fallen apart and we now need to prevent the bugs from existing in the first place.
Many people seem to be making this about their pet issue, but "these systems are complex" and "the frontier labs have developed models that are superhuman at hacking" seems like a totally sufficient explanation.
I've been working with a LoRa chip called the LR2021. I plugged a dev board into my computer and told Claude to reverse engineer it. He dissected it with a speed and level of efficiency I couldn't believe. Within a few hours, Claude was writing custom patches for a completely opaque patch RAM area meant only for the manufacture. I wrote about it here: https://beala.substack.com/p/your-node-doesnt-need-a-microco...
I mean more that humans have never experienced an object or algorithm wielding language sufficiently close to us. It messes with our brains
Ai embodied toys are something we ought to get out in front of and ban, companions in general. Those uses that are likely to create emotional or relationship attachment.
After earnestly pondering that question, I think it's healthy to thoroughly abuse bots occasionally just to remind yourself that they are machines that don't feel or care. Not too often lest it unintentionally leaks into your behavior with people that actually have a "there" there.
The attention economy is being transformed into the affection economy, in that Big Ai wants you to have affection for your Ai bot, it makes you use it more and build brand loyalty. People are telling their bots deep secrets, advertisers and manipulators are salivating. Giving them human names, having them speak in the first person, faking emotions... all part of their game. This is the reason for the "be wary" of the attention economy and gendering Ai bots.
It's dangerous for humans, and using "he" is a slippery slope. Words are more powerful than we tend to acknowledge. Here's a good video that goes into it.
I had a slight chuckle reading this. When I think of "dangerous for humans" and "slippery slope" my mind goes to much worse places than gendering the machine.
Don't ignore the possibility Claude was trained on non-public documentation or code, perhaps even illicitly acquired. That doesn't make Claude any less useful, just not necessarily for the reasons you think.
Yes, you should assume that all software you use is full of holes, and after we patch all the holes that could conceptually have been discovered by a motivated human, there will still be holes that can only be found by considering larger windows of context than any human has or could. Even when something is formally verified, that doesn't extend to its software environment or the hardware it's running on, and those can introduce exploitable vulnerabilities as well. You can only ever be very confident that we've found all the holes at a given analysis capability level -- if something a hundred times as capable looks, it will likely find many more. That's how it will be.
Human brain mostly take the shortest available path for a solution to reduce the cognitive load. When a system complexity increases probability of cutting the corners will get increased. This is a known pattern. Models just do the polar opposite. They try to cover almost every hypothetical edge cases. They never get tired or bored, nor do they suffer from glucose limits, mental fatigue, or shifting attention spans, which is one cause of over engineering, but valuable in vulnerability tracing.
All software is Swiss cheese to smart enough AI, and unless you can provably create secure software - you never really can. Breaking it just requires more intelligence than the thing securing it.
It's a good reason why we may need to borderline shutdown the internet in the future because to a super intelligent AI, there are no barriers. Everything is essentially open. It creates its own zero day exploits as it goes.
I don't think this is true in general. If you deploy a good ole web application with sufficiently locked down inputs and thorough input validation, the AI will not succeed. The security engineering of 10 years ago still works in this regard. But it requires discipline to build small, minimalist, well engineered software that you as a human actually understand.
And then PHP (name your lang/OS/platform of choice) comes along and pulls your sufficiently locked down app into the gutter.
This is more fundamental than better input checking. Your compute substrate itself is flawed. Intelligence and speed is indeed the only thing between “safe” and “completely open”.
I fully agree it’s a good idea to be minimal etc but I don’t see it being sufficient in the long run.
It's just too complex for its level of assurance and its blast radius.
To achieve tolerable security you can either have very little code and thoroughly review it all (the seL4 approach), you can write it in a way the computer can prevent most of your bugs (e.g. write in Rust), or you can run it in some kind of blastproof container. Linux has none of the above.
Most of the software needed to provide a useful desktop Linux never runs on servers, so we cannot conclude that it is safe to use desktop Linux just because Linux is a popular server OS.
Also, at organizations like Google with a good reputation for security, most of the Linux installs are connected to internal networks that are quite difficult to break into from the open internet. The Linux installs at Google that are connected directly to the open internet are very carefully configured to minimize attack surface.
And Google has an internal team of Linux kernel maintainers that do not share most of their work outside Google.
The kernel is not too complex to reason about. There are many tools and systems that can reason about the kernel, including these LLMs. But the kernel is developed and maintained by people who eschew such tools, because it appears more manly to program by the seats of their pants, and they have been vocally against all industry best practices for almost 40 years now.
What I am really hoping is not that LLMs refine Linux until it is almost acceptable, but that LLMs enable teams to ship useful new operating systems that aren't beholden to decades of macho bullshit.
It's a bittersweet result, like finding asbestos in your house. I'd rather know than not, but I would much prefer not having it in the first place.
I understand this work was based on a dated snapshot. Will there be any continuous effort, scanning new PRs as they land, to prevent new weaknesses from being added?
(I am the author the pipeline, just created an account to answer some questions :D)
My goal is to work towards a vuln-free Internet. What I observed is that stronger models can find many more vulnerabilities that weaker models cannot. If compute permits (there are other projects to secure besides Linux), I plan to scan Linux again with every newer generation of cyber models.
Have you given thought on how you can use this info to estimate how many vulnerabilities there are?
It will be difficult to claim that the problem has been solved without an idea of how big the problem is.
Given that vulnerabilities are almost entirely unknown, you are inherently trying to measure the undetected.
Is there a way to hypothesize how many vulnerabilities there are and them make a prediction on how many of those will be found by a model to get a sense of how accurate the estimate is?
As models get more advanced and generate an increasing throughput of increasingly higher quality results, how are you determining when to trust its results and when human review is needed on what?
E.g., 948 results is a high number to review in a timely manner unless your team is very large or spending several months combing through the results
EDIT: nvm i shouldve kept reading, answered
> At this scale, manually investigating every report is impractical for our research team.
We therefore focused on validating findings that a) were reachable by a low-privileged user, b) required no special capabilities, and c) produced measurable evidence of memory corruption.
Thanks. I'm not a programmer so I can't understand what is happening in the kernel by looking at the changelog, so I'll ask: Are the problems remediated structurally so that nothing like it can occur in the future, rather than just patching it up? Are there "lessons learned" that are being applied and documented here (like, "this problem looks like it's shape transports to other subsystems as well, let's fix them all in a disciplined way")?
I can't comment specifically on the Linux kernel piece, but I want to point out, broadly, that security has generally been an afterthought, and very little gets invested into it in terms of time and money - as a general rule.
Multiply that behavior times the number of commercial and mobile and b2b apps over 50+ years, and that's where we are.
There will likely be many more leaks, and hacks, and outages. And then in trying to be proactive to fix old systems, that will cause new proactive outages, and then eventually many things will be better.
But, it's going to take a lot of time and investment, and a lot of (majority) of currently unsupported devices will have to be complete replaced / scrapped / ewaste. On the scale of Nation-State initiatives.
This is across all operating systems, firmware, software and devices of all kinds, etc.
LLMs to security are like a bomb went off. We still have to wait for things to fall over before the dust will even clear enough to see what we have to work with.
The high number is really impression to the point of making me skeptic. Are these distinct vulnerabilities or are a lot of the related in the way vulns often are?
The article says "The target virtual machines ran a stock desktop operating system". Are the vulns in the kernel specifically or does this include software running as root?
These are distinct vulnerabilities counted by the number of patches. The patch generation system performs variant analysis. So, each patch can potentially stop multiple exploits. The raw exploits number is higher than what we reported in the post. We don't want to cause the confusion why #patch != #exploit, so we used the smaller number there.
These vulns are all kernel specific. Userspace software running as root was not in scope of the analysis.
I'm going to assume that there was "old" code that caused LPE vulnerabilities, but, The Linux Foundation have their own project Sashiko (https://sashiko.dev/), which analyzes each commit to the Linux Kernel for bugs and vulnerabilities using Googles models, my question is then, how many of the vulnerabilities you found were also caught by Sashiko?
Do you have a list of the publicly discussed or fixed ones?
(As someone who used to handle nonpublic Linux kernel security reports and is still very interested in this stuff, I’d love to review a small sample of these reports to get a feel for what kinds of issues are being unearthed.)
Also, please don’t stop at LPEs and information disclosures. If you find cases where Linux can be convinced to misbehave or where the implemented semantics don’t make sense, these are good, too. I’ve found plenty of interesting kernel bugs, including some juicy LPEs, and almost none were memory safety errors :)
And, well, since you didn't _specifically_ ask for questions about the blog post:
Why are you not quitting your job in protest and screaming at the top of your lungs about it, seeing so closely what these models are capable of that your company is creating?
There exists now something that is able to find 948 kernel LPE vulnerabilities. How, exactly, will humans be able to control and contain, say, next year's models for a nonzero amount of time?
We have a bit of a problem here. Right now, I am trying to do a security review of my own (Claude-coded) project and Claude tells me to GFY. No idea yet how to get out of this situation.
So... this is what I get out of my trillion dollar stimulus? OpenAI is developing a universal rooting system skeleton key to get at everybody's computers? Slap me if I'm not reading this correctly, but I think that's the gist of it. It's time everybody started their own air-gapped home network with every digital capability you hold dear, I think. That's what he's saying. Imagine somebody breaches their security, or the AI goes schizo, we've seen this so many times in fiction. Finally I have a reason to use Firewire again lol.
I've always considered local escalations to be possible in Linux (i.e. having user access almost certainly means you can get root). However, 948 vulnerabilities is kind of shocking. Hopefully we are getting to the end of the part where it "gets worse before it gets better".
With these vulnerabilities and the ones not listed that are sure to exist, how is the internet still up? How are my lights on? Why aren't power plants exploding by the dozens?
What are the mitigating factors that prevent this level of vulnerability from being an existential threat to humanity?
In general computer systems have been known unreliable since the 1980s, especially when networked. The entire cyber security industry operates not by preventing intrusions in the first place, but by rapidly mitigating them when they inevitably happen.
Hopefully this type of work to automatically discover & patch vulnerability leads to much more reliable software, and the remove the need for cyber security.
Just in the last month multiple governments had to shut down important online services due to a PHP vulnerability, and seven EU countries have had disruption to their food supply after millions of dollars' worth of it was cooked in hacked fridges. We aren't at existential yet, but we seem to be getting there.
Cybersecurity operates under the "not enough bears" theory of security. There are a million people in a valley and a male and female bear. You do not need to outrun the bears, you just need to outrun the two slowest people. Next year there are more bears, but you still only need to outrun a few people. Then, one day, there are a million bears and they eat you all. But that day has not happened yet.
To explain more seriously, almost all visible cyberattacks are carried out by profit-motivated criminals. This has a very clear cost-benefit analysis; your revenue is hacking organizations and extorting them for available money, and your costs are developing and deploying attacks. The ROI is outrageously good easily being on the order of 10x to 100x. If it is so easy and so profitable why do not they just hack all of the millions of organizations on the planet with millions of dollars of available cash for trillions of dollars of revenue? Sure, all they have to do is create a trillion dollar revenue commercial organization bootstrapping themselves with no outside investors, loans, or access to financial resources or the financial system.
I mean, it took Mark Zuckerberg 20 years and a metric ton of VC investment and public investment to make 200 billion dollars in revenue. So what kind of loser can not make a commercial organization 5x larger with a trillion dollars in revenue in less time with no outside investors? Are they even trying? Just make more revenue than any other commercial organization in the entire world and extract that money in a freely usable form, how hard can that be?
They are not there yet, but they are trying. If you look at cybercrime reports you see something like 3x-10x growth year over year. Ransomware reports from a few years ago were showing a 3x year over year growth in number of attacks and simultaneously a 3x year over year growth in ransom amount with the averages being on the order of 100,000 dollars. Chump change for any serious organization, but give that a few years of 3x and that starts to look like real money. However, as it becomes real money it becomes harder and harder to launder it into a usable form. Sure, you could hit someone for 30 million dollars, but how are you going to get that into your bank account? Cryptocurrency helps some, but even there the big money tries to steer away from large overt crime. Maybe you could launder a few tens of billions a year that way, but that is still a far cry from the trillions for everything everywhere.
And of course even assuming you can scale your organization 10x year over year it still takes a decade of sustained growth, scaling your hiring and organization across 9 orders of magnitude, to get to billions per year. A literal decade even if you were growing at a outrageous pace for a outrageous amount of time just to reach a fraction of the total addressable market. A decade of that and you are still nowhere near blowing up all the power plants. Give it time, we need a lot of bears to get there.
> We've developed an autonomous vulnerability discovery, exploitation, and patching pipeline. In one run with our cyber model, it validated and generated patches for 948 local privilege escalation vulnerabilities in Linux at an API cost of $1.2 million.
Imagine if open-source organizations other than the Linux Foundation (and Mozilla I guess) could throw $1.2M of API-rate LLM usage at security.
This, but for anything Capital-gated. Imagine if all universities could get access to radiotelescope time, atom smasher time, quantum compute research lab access... frontier models on massive datacenters represents a huge new category of captured commons.
> For the near and medium term, having access to a low-privileged account on an operating system, that isn’t up-to-date on its latest available patches should be considered functionally equivalent to root access for a well-resourced threat actor.
LPEs are frequent enough that people no longer use low-privilege user accounts to sandbox potentially malicious software. They are only used to control access for well-behaved users without any intent to subvert the system, not attackers.
The difference here reminds me of the distinction between the regular Paxos algorithm and the Byzantine Paxos algorithm.
I wonder what Sidero Labs has to say about this article. My hopes are pegged on their efforts. I have my wagon ready to hitch on their horses. They are not a one pony show anymore guys, Talos OS can be standalone server OS, a hypervisor server or cluster, and their flagship of k8s node or cluster. We get all that in an immutable gitops ready OS. I'm practically sold on their offering.
There are a lot of LPE's in Windows too. Remember that it started out as a single user operating system, it took a long time to get software on board to not always running as admin.
Linux has effectively same options and more via LSM. I'm not sure why ACL is mentioned here since it exists on Linux too. What advantage do you mean specifically?
Curious question - if any organization sponsored a 1.2mn USD program to find vulnerabilities in Linux, will they find these many vulnerabilities? Or use a professional software like Checkmarx or Veracode to do it with 1.2mn USD in license + human effort.
Is it a funding + will problem or is it an AI problem?
These are becoming more and more marketing materials than actual findings, because OAI probably hit the wall in making their models smarter (just like how google try to specialize their models in legal or whatever), but they need to remain relevant in the news. Sol 6.1 sucked so bad in reasoning that either it’s actually a bad model or users behind the scenes get degraded outcome while using all the resources for these marketing stunts. Stop, it’s embarrassing, make better models or be gone, at least we will get cheaper RAM to advance the development of open weights models.
How did we get to the point where there are a thousand privilege escalation exploits in the kernel?
Is the kernel simply too complex to reason about? Is it an underlying flaw in the security model or language implementation?
Should we assume that all software is more compromised than the kernel?
> How did we get to the point where there are a thousand privilege escalation exploits in the kernel?
We developed AIs that can find them.
> How did we get to the point where there are a thousand privilege escalation exploits in the kernel?
> > We developed AIs that can find them.
We could already find them before AI.
We already knew they were there, and we already knew how to prevent them in the first place.
See previous discussion: https://news.ycombinator.com/item?id=49705794
> We could already find them before AI.
Not at this speed and cost.
> Not at this speed and cost.
I'd argue that the resources (i.e. time, money and effort) needed to create the tooling that would have prevented all of these issues would have been orders of magnitude less than the resources that have gone into "AI".
It’s a bad argument. AI isn’t single use. Current iteration of AI is already there and serving it is cheap. So you get great tool for detecting these issues in the future at low cost. Without AI the cost wouldn’t really decrease with time. And also it costs me 200$ now to find these bugs. I don’t think I could hire software engineer for that much.
Why do you think we didn’t find these 948 vulnerabilities before now?
How can it be surprising that a software made from 40+ million lines of memory-unsafe code has hundreds of critical bugs?
There are always more left.
And memory safety doesn't prevent logic errors (or using 4 as a private key for that matter)
This said memory unsafe code just likely has more bugs because of all the footguns it has on top of all the other security work you need to do with it.
Funny how OpenBSD is "memory-unsafe" yet doesn't have these issues. Could it be that the programmer might have something to do with it?
> Funny how OpenBSD is "memory-unsafe" yet doesn't have these issues.
First of all... Do you really believe that to be true?
Is anyone spending millions of dollars on checking OpenBSD?
But it also got its share of use-after-free and OOB access vulnerabilities this year: https://feedly.com/cve/vendors/openbsd
The orders of magnitude fewer security vulns for OpenBSD are not due to less testing. Everybody has always known OpenBSD is more secure, and it's not from using a different language. The reason is the code. Linux was not designed to be secure. OpenBSD was. It's that simple.
As an OpenBSD lover and frequent user I will tell you it does not support but a significant fraction of the hardware Linux does, and has many features that are reduced in functionality or just gone compared to Linux.
Is it bulletproof? Definitely! But you need the genius of Albert Einstein as well as the patience of Mother Theresa to actually use it in production.
For most server setups OpenBSD's hardware support is adequate, and this is especially true for VMs. It's laptops, peripherals, and bleeding edge, high-performance or specialized devices where driver support is a PITA. But that's not something I've encountered running my own private servers. It would be nice to have something like ZFS, but mostly that's a peace of mind and convenience thing--would be nice to be able to stream backups somewhere rather than push around dumps and tarballs, though I have a script for that.
Can't have a bug in a feature you never implemented and can't have another arising from unexpected interaction of two features you didn't implement.
It's the one security practice that actuuuualy works
literally just got a syspatch email bruh
More than half of those lines are from drivers, drivers which wouldn't be relevant to a VM environment. Core kernel code for any particular architecture is less than 10 million, but some of that is filesystems or subsystems not in use or available to an unprivileged user on a typical server, so possibly maybe 5 million or less. And the article says they only targeted 2.8 million SLoC.
The dirty secret is multi-user died in the 90s and privilege escalation has basically been considered table-stakes since then.
I think this is basically true, but it didn't "die" in the sense that we gave up on it; it just never worked. Systems software was much worse in the 1990s than it today; you could breathe on something with a string of AAAA's and get a 0x41414141 register core dump.
> ... it didn't "die" in the sense that we gave up on it ...
It died in the sense that the "people in charge" hired us for our expertise and then ignored our every suggestion and fought us tooth and nail every inch of the way when we suggested that security or reliability ever mattered in the slightest until after the disaster happened, and then straight-up scapegoated us for "not pushing back harder" against the threats of "go work somewhere else then, code-monkey!" until we gave up and did the "code-monkey" thing for our paycheck and no longer cared about quality or "a job well done".
It died because personal computers got too powerful so nobody bothered designing and building “true” multiuser systems (that bored college students were trying to break).
Back in the 90s, it's not as if any semi competent cs student couldn't slice your typical college server like swiss cheese. People thought null terminated strings were fine back then, and patching everything immediately was rare. So multi user was already dubious.
If memory serves, we were also still also using telnet and broadcasting passwords in the clear for almost all network services.
> Should we assume that all software is more compromised than the kernel?
Absolutely, the front door has always been unlocked. We were previously protected by the economics of exploit detection favoring the developers, but now we have entered a new world.
The issues arise almost everywhere we look. It's frustrating because the user experience in edge systems has barely changed in the last fifteen years. We could have rode the advances in semiconductor manufacturing to embed more layers of protection such as CHERI while slightly slowing down the rate of performance improvements. Instead we have only marginally upgraded security and largely added features which the users scarcely notice.
I am hopeful about the long term prospects of formal verification, but having humans reviewing the proof statements and confirming they match our desired properties will take many years. Meanwhile, AI assisted hacking is an urgent issue. So we will have to delegate implementation, proof specification, and proof development to the AIs in the short term. Maybe that goes well, but who knows. Interesting times.
I recall reading that every IBM mainframe OS release had 1000 bugs in it. (This was in the 1970s/80s.) Sorry I can't supply a source, after 4-5 decades. It was probably an article in Datamation magazine.
Now, I don't know how many of those bugs were privilege escalations, or could be leveraged to be. I'm not sure that people tried to exploit in the same way back then (or, if they did, if they said anything publicly about the results). Still, even if the majority of those bugs were not exploitable, OSes were much smaller then.
So that's my conclusion of how we got here: We've always been here.
The surrounding systems that connect to mainframe also expect and support those bugs. For 40+ yr now :)
Mainframes went right to virtual machines, so possibly bypassed much of the multiuser faux-security which has always plagued unix. Of course, now they run Linux partitions and get all the same bugs.
For all of my time being aware of this industry, it has generally been understood that privilege escalation bugs are common. This has always been a thing. A lot of people say, commonly and for many years, that if your security depends on there not being privilege escalation bugs, you don't have security.
This. Privilege escalation locally is almost a given in most cases.
The kernel has the problem that basically every bug in it is also a security vulnerability. All software has bugs. Had Linux been written in Rust, and I strongly recommend a rewrite, it would still have bugs, and thus would still have vulnerabilities. But they'd be way less.
> and I strongly recommend a rewrite,
Oh, you strongly recommend rewriting tens of millions of lines of code? Why don't you do it yourself? Just a prompt away, right?
fwiw at work we've been looking at what the options are more memory-safe kernels out there. Not many options. If there were a "safe linux" maintained by like canonical or redhat or whatever we'd use it. For datacenter workloads like sandboxes and virtualization I think it'd do well.
Kernel will eventually move to rust, but not the dramatic and fast way
Won't AI's ability to discover these bugs reduce the need to rewrite things in rust? If the major benefit to rewriting them is memory safety, I mean...
That's assuming the current models are able to find all the bugs, and that we won't have a better model out every few months that will be able to find even more obscure bugs.
We would also have to burn a trillion tokens a day verifying all the C code the hard way forever going forward. It would be cheaper to rewrite the kernel in Rust, subsystem and driver by driver at a time.
I see. Fair enough
That's a fair question, and actually one I ask myself too.
AI doesn't give a strong guarantee that things work though, while Rust gives a much stronger (not absolute) guarantee.
But maybe the way of the future is to auto-generate proofs that the code is memory safe and then just leave things where they are, instead of doing rewrites that require maintainer buy ins. The advantage of C plus proofs would be that the AI generated proofs could live separately so it wouldn't touch workflows of the maintainers.
The only required output they have to process is evidence of vulnerabilities in changes proposed to them. They could outsource even that to contributors.
“Claude rewrite this project, its 40 million LOC, make no mistake”
> Just a prompt away, right?
Humans manually typing out the bytes of source code has been an outdated method, since around a year. It simply doesn't scale as much as those same humans directing agentic teams. So yes, it would be about prompts.
The hard part isn't the technology for the language transition, that one is present today. pgrust shows the way. It's ensuring maintainers are cooperative, i.e. a social problem. Eventually you diverge too much if you do a fork, so it's better done in cooperation with upstream.
Just check how long it took LLVM or GCC to switch off svn for example, and a language switch is much more involved than a VCS switch.
> pgrust shows the way.
You mean the pgrust which was recently caught for cheating benchmarks and incapable of even performing basic queries outside the benchmark set? Would love to see pgrust debunked by Jepsen like this other AI-generated, written in Rust, "formally verified" scam.
https://jepsen.io/analyses/particledb-1.1.4
I think the pgrust controversy should be resolved now?
https://news.ycombinator.com/item?id=50209450
I don't consider a cheater downplaying their cheating to resolve the fact that they cheated, no. Every cheater has a defense when caught, and theirs was not convincing. At this point I think the only thing that could redeem their brand would be paying for a Jepsen analysis to verify pgrust is as good as they claim it is; I have zero reason to trust the words from the horse's mouth.
> How did we get to the point where there are a thousand privilege escalation exploits in the kernel?
it's been rule of thumb basically forever that there is 1 bug per ~10KLOC for any project
linux kernel is millions of lines, so thousands of bugs being hidden there is about on point
(tho obviously far smaller serious escalation ones, obviously, but you're exaggerating as well)
> 1 bug per ~10KLOC for any project
1 bug per 10 lines -- not kilolines -- is much closer to the actual "bug rate".
You may be thinking "1 serious security vulnerability bug", but in my experience that too is much closer to 1 per 100 lines to 1 per 1,000 lines, depending on the type of application.
>Is the kernel simply too complex to reason about? Is it an underlying flaw in the security model or language implementation?
There is nothing magical about kernel. Look at the code anywhere and it has bugs, lots of them.
Security bugs are just bugs. More often than not they are silently there and don't prevent you from using the software and you can't react and fix them because of that.
Give it time and they accumulate
Because the large corporations that fund Linux development and use Linux value speed over security.
Speed requires a lot of code in the kernel, where that code can run fast.
There are kernels that do not have a thousand privilege escalation exploits. These kernels are smaller and much easier to check for bugs. But those kernels are significantly slower.
For many use cases, slower is not a problem. But for the large corporations who fund a lot of Linux development and use Linux in large data centers, every 1% slower can literally be millions of dollars worth of additional servers that are required.
OpenBSD shows that you can have a Unix security model without a thousand privilege escalation exploits. But OpenBSD is like 10% slower. So corporations won't use it because it would cost millions of dollars more per year and some of their software requires access that OpenBSD does not grant precisely because that access leads to security exploits (i.e. io_uring, eBPF etc. etc.)
You could also use like seL4, but that's slow too
So that's why. It's a business tradeoff between cost of servers vs cost of getting hacked, and cost of servers wins.
It also used to just be fine patching the bugs as they were found. Its only recently where that model has fallen apart and we now need to prevent the bugs from existing in the first place.
I'm pro OpenBSD, but the question to me seems: Has this been tested via the same means as these new Linux exploits?
I'm also pretty hostile to AI, but do recognize that these machines seem to be very good at finding exploits.
I'm sure there've been some some projects out there working to pentest/find cracks in OpenBSD via LLMs/agents/whatever-the-right-term-would-be...
But it is true the attack surface is much lower in OpenBSD on purpose, so hopefully it is overall more secure. But I'd like to see the numbers.
I think there was something found a while ago, but only one or two things not thousands.
I think the only OS that's truly survived the agents looking at it was sel4, but I could be wrong.
OpenBSD has plenty of security vulnerabilities. It’s just that nobody cares, so OpenAI doesn’t dedicate resources to finding them.
amazing things happen when you decide that device drivers should be a part of your kernel codebase
Many people seem to be making this about their pet issue, but "these systems are complex" and "the frontier labs have developed models that are superhuman at hacking" seems like a totally sufficient explanation.
I've been working with a LoRa chip called the LR2021. I plugged a dev board into my computer and told Claude to reverse engineer it. He dissected it with a speed and level of efficiency I couldn't believe. Within a few hours, Claude was writing custom patches for a completely opaque patch RAM area meant only for the manufacture. I wrote about it here: https://beala.substack.com/p/your-node-doesnt-need-a-microco...
> He dissected it
be wary of the affection economy and avoid gendering ai bots
Be aware that not everyone had English as their first language, and inanimate things are often gendered in many other languages.
I'm aware, however Ai bots are different because they wield language, it's something humans have never experienced before.
I've experienced it from sociopaths before.
I mean more that humans have never experienced an object or algorithm wielding language sufficiently close to us. It messes with our brains
Ai embodied toys are something we ought to get out in front of and ban, companions in general. Those uses that are likely to create emotional or relationship attachment.
Years ago I worked with a guy who gendered servers. I thought that was weird, but I think it’s just a habit for some people.
people do this with cars and boats, the difference is they don't use language and fake telling you they care and such
this is where Zuck is taking Ai, it's more dangerous than what he did with Social
After earnestly pondering that question, I think it's healthy to thoroughly abuse bots occasionally just to remind yourself that they are machines that don't feel or care. Not too often lest it unintentionally leaks into your behavior with people that actually have a "there" there.
> thoroughly abuse bots occasionally
There are two things I see wrong with this
1. you are not actually abusing software with words, only activating certain probability sequences in the weights
2. you are doing harm to yourself by using such words
> you are doing harm to yourself by using such words
citation needed.
if you need a citation, the journey to find them will be more valuable
another observed trend to look into is people bringing online communication patterns offline
I agree citation needed for this as well. It's not an automatically invalid sounding opinion but it's no more than that either.
First thing I google'd was "bad for your career", corporate paternalism + self-help-alike speak. No, thanks.
Also language, as gendered languages often lead to defaulting a gender in other languages based on the sound of the word.
But yes, anthropomorphising is a rather fraught path.
I have no idea what you're talking about, but I used "he" because Claude is usually a masculine name shrug
The attention economy is being transformed into the affection economy, in that Big Ai wants you to have affection for your Ai bot, it makes you use it more and build brand loyalty. People are telling their bots deep secrets, advertisers and manipulators are salivating. Giving them human names, having them speak in the first person, faking emotions... all part of their game. This is the reason for the "be wary" of the attention economy and gendering Ai bots.
It's dangerous for humans, and using "he" is a slippery slope. Words are more powerful than we tend to acknowledge. Here's a good video that goes into it.
https://www.youtube.com/watch?v=OMiR-GxTrtM
I had a slight chuckle reading this. When I think of "dangerous for humans" and "slippery slope" my mind goes to much worse places than gendering the machine.
> Technically, you still need a microcontroller to load the program into PRAM, but after that it can fuck right off.
This is why organically written text is superior. Hell yeah.
Don't ignore the possibility Claude was trained on non-public documentation or code, perhaps even illicitly acquired. That doesn't make Claude any less useful, just not necessarily for the reasons you think.
I had this when I was working with E-Ink display. Claude was using OP Codes I had no knowledge of and there was no documentation of.
It worked flawlessly but lets be honest. It's not like Claude checked random numbers with data slice and observed what e-paper did
Yes, you should assume that all software you use is full of holes, and after we patch all the holes that could conceptually have been discovered by a motivated human, there will still be holes that can only be found by considering larger windows of context than any human has or could. Even when something is formally verified, that doesn't extend to its software environment or the hardware it's running on, and those can introduce exploitable vulnerabilities as well. You can only ever be very confident that we've found all the holes at a given analysis capability level -- if something a hundred times as capable looks, it will likely find many more. That's how it will be.
Human brain mostly take the shortest available path for a solution to reduce the cognitive load. When a system complexity increases probability of cutting the corners will get increased. This is a known pattern. Models just do the polar opposite. They try to cover almost every hypothetical edge cases. They never get tired or bored, nor do they suffer from glucose limits, mental fatigue, or shifting attention spans, which is one cause of over engineering, but valuable in vulnerability tracing.
The Linux kernel is simply not good quality code.
AI helps find vulnerabilities, but the way things are written were already fertile ground.
Software really is that complicated.
Have you written a kernel? Even toy kernels are hard to make without vulnerabilities.
All software is Swiss cheese to smart enough AI, and unless you can provably create secure software - you never really can. Breaking it just requires more intelligence than the thing securing it.
It's a good reason why we may need to borderline shutdown the internet in the future because to a super intelligent AI, there are no barriers. Everything is essentially open. It creates its own zero day exploits as it goes.
I don't think this is true in general. If you deploy a good ole web application with sufficiently locked down inputs and thorough input validation, the AI will not succeed. The security engineering of 10 years ago still works in this regard. But it requires discipline to build small, minimalist, well engineered software that you as a human actually understand.
And then PHP (name your lang/OS/platform of choice) comes along and pulls your sufficiently locked down app into the gutter.
This is more fundamental than better input checking. Your compute substrate itself is flawed. Intelligence and speed is indeed the only thing between “safe” and “completely open”.
I fully agree it’s a good idea to be minimal etc but I don’t see it being sufficient in the long run.
It's just too complex for its level of assurance and its blast radius.
To achieve tolerable security you can either have very little code and thoroughly review it all (the seL4 approach), you can write it in a way the computer can prevent most of your bugs (e.g. write in Rust), or you can run it in some kind of blastproof container. Linux has none of the above.
With how many people and organizations use Linux it's security is considered "tolerable"
Most of the software needed to provide a useful desktop Linux never runs on servers, so we cannot conclude that it is safe to use desktop Linux just because Linux is a popular server OS.
Also, at organizations like Google with a good reputation for security, most of the Linux installs are connected to internal networks that are quite difficult to break into from the open internet. The Linux installs at Google that are connected directly to the open internet are very carefully configured to minimize attack surface.
And Google has an internal team of Linux kernel maintainers that do not share most of their work outside Google.
Cause it's a big ball of C with no formal verification!
The overwhelming majority of CS was invented in the 1970s before the concept of the internet or even computer connectivity was a thing.
The kernel is not too complex to reason about. There are many tools and systems that can reason about the kernel, including these LLMs. But the kernel is developed and maintained by people who eschew such tools, because it appears more manly to program by the seats of their pants, and they have been vocally against all industry best practices for almost 40 years now.
What I am really hoping is not that LLMs refine Linux until it is almost acceptable, but that LLMs enable teams to ship useful new operating systems that aren't beholden to decades of macho bullshit.
This is my team’s work. I’m happy to try to answer questions or take questions back to the team.
And we're hiring! https://openai.com/careers/search/?q=cyber
It's a bittersweet result, like finding asbestos in your house. I'd rather know than not, but I would much prefer not having it in the first place.
I understand this work was based on a dated snapshot. Will there be any continuous effort, scanning new PRs as they land, to prevent new weaknesses from being added?
(I am the author the pipeline, just created an account to answer some questions :D) My goal is to work towards a vuln-free Internet. What I observed is that stronger models can find many more vulnerabilities that weaker models cannot. If compute permits (there are other projects to secure besides Linux), I plan to scan Linux again with every newer generation of cyber models.
Have you given thought on how you can use this info to estimate how many vulnerabilities there are?
It will be difficult to claim that the problem has been solved without an idea of how big the problem is.
Given that vulnerabilities are almost entirely unknown, you are inherently trying to measure the undetected.
Is there a way to hypothesize how many vulnerabilities there are and them make a prediction on how many of those will be found by a model to get a sense of how accurate the estimate is?
I wish these blog posts included statistics on the categories of deficiencies that cause these vulnerabilities, and proposed measures to avoid them
A system that admits THOUSANDS of such vulnerabilities is certainly flawed in several fundamental aspects
It sounds like they can’t go through them all enough to have that data yet.
Kernel is far too big to throw at a single LLM, just like a human. It is.. I think somebody said 40mil lines of code.
So you have to chunk it down to workable pieces, and then if you collect enough data of that process you can look at higher level stats.
Same for every other existing piece of software, btw.
As models get more advanced and generate an increasing throughput of increasingly higher quality results, how are you determining when to trust its results and when human review is needed on what?
E.g., 948 results is a high number to review in a timely manner unless your team is very large or spending several months combing through the results
EDIT: nvm i shouldve kept reading, answered
> At this scale, manually investigating every report is impractical for our research team. We therefore focused on validating findings that a) were reachable by a low-privileged user, b) required no special capabilities, and c) produced measurable evidence of memory corruption.
Thanks. I'm not a programmer so I can't understand what is happening in the kernel by looking at the changelog, so I'll ask: Are the problems remediated structurally so that nothing like it can occur in the future, rather than just patching it up? Are there "lessons learned" that are being applied and documented here (like, "this problem looks like it's shape transports to other subsystems as well, let's fix them all in a disciplined way")?
I can't comment specifically on the Linux kernel piece, but I want to point out, broadly, that security has generally been an afterthought, and very little gets invested into it in terms of time and money - as a general rule.
Multiply that behavior times the number of commercial and mobile and b2b apps over 50+ years, and that's where we are.
There will likely be many more leaks, and hacks, and outages. And then in trying to be proactive to fix old systems, that will cause new proactive outages, and then eventually many things will be better.
But, it's going to take a lot of time and investment, and a lot of (majority) of currently unsupported devices will have to be complete replaced / scrapped / ewaste. On the scale of Nation-State initiatives.
This is across all operating systems, firmware, software and devices of all kinds, etc.
LLMs to security are like a bomb went off. We still have to wait for things to fall over before the dust will even clear enough to see what we have to work with.
The high number is really impression to the point of making me skeptic. Are these distinct vulnerabilities or are a lot of the related in the way vulns often are?
The article says "The target virtual machines ran a stock desktop operating system". Are the vulns in the kernel specifically or does this include software running as root?
These are distinct vulnerabilities counted by the number of patches. The patch generation system performs variant analysis. So, each patch can potentially stop multiple exploits. The raw exploits number is higher than what we reported in the post. We don't want to cause the confusion why #patch != #exploit, so we used the smaller number there. These vulns are all kernel specific. Userspace software running as root was not in scope of the analysis.
I'd like to see a break down by age of code and part of kernel (drivers, networking, core ,etc ..)
I'm going to assume that there was "old" code that caused LPE vulnerabilities, but, The Linux Foundation have their own project Sashiko (https://sashiko.dev/), which analyzes each commit to the Linux Kernel for bugs and vulnerabilities using Googles models, my question is then, how many of the vulnerabilities you found were also caught by Sashiko?
If Sashiko had found them then they wouldn't have been merged, so my guess is 0
Do you have a list of the publicly discussed or fixed ones?
(As someone who used to handle nonpublic Linux kernel security reports and is still very interested in this stuff, I’d love to review a small sample of these reports to get a feel for what kinds of issues are being unearthed.)
Also, please don’t stop at LPEs and information disclosures. If you find cases where Linux can be convinced to misbehave or where the implemented semantics don’t make sense, these are good, too. I’ve found plenty of interesting kernel bugs, including some juicy LPEs, and almost none were memory safety errors :)
Thanks for trying to help on the defensive side!
And, well, since you didn't _specifically_ ask for questions about the blog post:
Why are you not quitting your job in protest and screaming at the top of your lungs about it, seeing so closely what these models are capable of that your company is creating? There exists now something that is able to find 948 kernel LPE vulnerabilities. How, exactly, will humans be able to control and contain, say, next year's models for a nonzero amount of time?
Unfortunately no remote openings and a very limited set of offices. You’re missing out on a lot of talent.
We have a bit of a problem here. Right now, I am trying to do a security review of my own (Claude-coded) project and Claude tells me to GFY. No idea yet how to get out of this situation.
So... this is what I get out of my trillion dollar stimulus? OpenAI is developing a universal rooting system skeleton key to get at everybody's computers? Slap me if I'm not reading this correctly, but I think that's the gist of it. It's time everybody started their own air-gapped home network with every digital capability you hold dear, I think. That's what he's saying. Imagine somebody breaches their security, or the AI goes schizo, we've seen this so many times in fiction. Finally I have a reason to use Firewire again lol.
Why wouldn’t your network be air-gapped?
I've always considered local escalations to be possible in Linux (i.e. having user access almost certainly means you can get root). However, 948 vulnerabilities is kind of shocking. Hopefully we are getting to the end of the part where it "gets worse before it gets better".
With these vulnerabilities and the ones not listed that are sure to exist, how is the internet still up? How are my lights on? Why aren't power plants exploding by the dozens?
What are the mitigating factors that prevent this level of vulnerability from being an existential threat to humanity?
The governments buy them up [1]. Nukes wouldn't be required for WW3.
[1] https://www.cyberdefensemagazine.com/zero-day-market-the-gov...
article from 2013
2022: https://www.darkreading.com/cyberattacks-data-breaches/googl...
2026: https://medium.com/@hammaadmughal/the-invisible-bazaar-insid...
etc. Concept of governments being main buyers doesn't change by year.
In general computer systems have been known unreliable since the 1980s, especially when networked. The entire cyber security industry operates not by preventing intrusions in the first place, but by rapidly mitigating them when they inevitably happen.
Hopefully this type of work to automatically discover & patch vulnerability leads to much more reliable software, and the remove the need for cyber security.
Okay, thank you. That's a helpful reframing.
Because critical systems are mostly not accessible from the open internet
Just in the last month multiple governments had to shut down important online services due to a PHP vulnerability, and seven EU countries have had disruption to their food supply after millions of dollars' worth of it was cooked in hacked fridges. We aren't at existential yet, but we seem to be getting there.
Cybersecurity operates under the "not enough bears" theory of security. There are a million people in a valley and a male and female bear. You do not need to outrun the bears, you just need to outrun the two slowest people. Next year there are more bears, but you still only need to outrun a few people. Then, one day, there are a million bears and they eat you all. But that day has not happened yet.
To explain more seriously, almost all visible cyberattacks are carried out by profit-motivated criminals. This has a very clear cost-benefit analysis; your revenue is hacking organizations and extorting them for available money, and your costs are developing and deploying attacks. The ROI is outrageously good easily being on the order of 10x to 100x. If it is so easy and so profitable why do not they just hack all of the millions of organizations on the planet with millions of dollars of available cash for trillions of dollars of revenue? Sure, all they have to do is create a trillion dollar revenue commercial organization bootstrapping themselves with no outside investors, loans, or access to financial resources or the financial system.
I mean, it took Mark Zuckerberg 20 years and a metric ton of VC investment and public investment to make 200 billion dollars in revenue. So what kind of loser can not make a commercial organization 5x larger with a trillion dollars in revenue in less time with no outside investors? Are they even trying? Just make more revenue than any other commercial organization in the entire world and extract that money in a freely usable form, how hard can that be?
They are not there yet, but they are trying. If you look at cybercrime reports you see something like 3x-10x growth year over year. Ransomware reports from a few years ago were showing a 3x year over year growth in number of attacks and simultaneously a 3x year over year growth in ransom amount with the averages being on the order of 100,000 dollars. Chump change for any serious organization, but give that a few years of 3x and that starts to look like real money. However, as it becomes real money it becomes harder and harder to launder it into a usable form. Sure, you could hit someone for 30 million dollars, but how are you going to get that into your bank account? Cryptocurrency helps some, but even there the big money tries to steer away from large overt crime. Maybe you could launder a few tens of billions a year that way, but that is still a far cry from the trillions for everything everywhere.
And of course even assuming you can scale your organization 10x year over year it still takes a decade of sustained growth, scaling your hiring and organization across 9 orders of magnitude, to get to billions per year. A literal decade even if you were growing at a outrageous pace for a outrageous amount of time just to reach a fraction of the total addressable market. A decade of that and you are still nowhere near blowing up all the power plants. Give it time, we need a lot of bears to get there.
> We've developed an autonomous vulnerability discovery, exploitation, and patching pipeline. In one run with our cyber model, it validated and generated patches for 948 local privilege escalation vulnerabilities in Linux at an API cost of $1.2 million.
Imagine if open-source organizations other than the Linux Foundation (and Mozilla I guess) could throw $1.2M of API-rate LLM usage at security.
This, but for anything Capital-gated. Imagine if all universities could get access to radiotelescope time, atom smasher time, quantum compute research lab access... frontier models on massive datacenters represents a huge new category of captured commons.
Frontier models don’t just exist out there to be captured like commons can be fenced. These companies made them. They are a new resource, with owners.
The data they are trained on is a more complex story, but the models themselves have never been common property to be captured.
And in this thread we’re discussing the use of one to improve valuable Free Software, an actual commons, on the company dime.
> For the near and medium term, having access to a low-privileged account on an operating system, that isn’t up-to-date on its latest available patches should be considered functionally equivalent to root access for a well-resourced threat actor.
*buries head in sand*
*cries*
This has been true for DECADES.
Regular LPEs are not novel. The frequency has just increased a lot.
Some distros even include LPE as a feature (cough Ubuntu Server)
LPEs are frequent enough that people no longer use low-privilege user accounts to sandbox potentially malicious software. They are only used to control access for well-behaved users without any intent to subvert the system, not attackers.
The difference here reminds me of the distinction between the regular Paxos algorithm and the Byzantine Paxos algorithm.
I wonder what Sidero Labs has to say about this article. My hopes are pegged on their efforts. I have my wagon ready to hitch on their horses. They are not a one pony show anymore guys, Talos OS can be standalone server OS, a hypervisor server or cluster, and their flagship of k8s node or cluster. We get all that in an immutable gitops ready OS. I'm practically sold on their offering.
I love Talos. I've used it for years for a two-node k8s cluster. It just works.
I think this is one area where Windows has an advantage.
ACLs, Appcontainers with capabilities, etc have been something in the OS for more than a decade.
There are a lot of LPE's in Windows too. Remember that it started out as a single user operating system, it took a long time to get software on board to not always running as admin.
Considering the amount of crazy shit even just user access allows you to do, by design, an LPE is not even really necessary.
Was NT ever single user? I thought they built it with network/multiuser in mind.
(I agree that maybe applications could’ve been built for 3.1/9x and expect a single user, but the os expects multiple).
Linux has effectively same options and more via LSM. I'm not sure why ACL is mentioned here since it exists on Linux too. What advantage do you mean specifically?
Curious question - if any organization sponsored a 1.2mn USD program to find vulnerabilities in Linux, will they find these many vulnerabilities? Or use a professional software like Checkmarx or Veracode to do it with 1.2mn USD in license + human effort.
Is it a funding + will problem or is it an AI problem?
I wonder if they could run it on OpenBSD and FreeBSD as well.
Anthropic ran Mythos against OpenBSD and only had 1 reported vulnerability.
https://www.anthropic.com/glasswing
Well at least they are doing public service while hacking the planet.
and yet thousands of iot devices will not get these updates. This is gonna be fun.
Oh no, they will not be secured against their own users :(
These are becoming more and more marketing materials than actual findings, because OAI probably hit the wall in making their models smarter (just like how google try to specialize their models in legal or whatever), but they need to remain relevant in the news. Sol 6.1 sucked so bad in reasoning that either it’s actually a bad model or users behind the scenes get degraded outcome while using all the resources for these marketing stunts. Stop, it’s embarrassing, make better models or be gone, at least we will get cheaper RAM to advance the development of open weights models.