korg.sol retweeted
I built 𝚌𝚙𝚞-𝚌𝚑𝚊𝚝, a reproduction of @openai's findings of covert communication using the CPU.
It runs on a generic host with 2+ Firecracker microVMs with no internet access. Works on x86 and even my M5 chip.
I was able to chat via 3 separate mechanisms:
① CPU contention
② Cache eviction timing
③ DRAM bandwidth contention
… across multiple CPUs, defeating all Firecracker's production recommendations (no SMT, no KSM, tenants pinned to different cores.)
Essentially, if shared hardware is involved, there's a good chance your agents can communicate covertly. And with really decent latency.
The good news: this is decidedly different from a *side-channel attack* where the agents could steal information (Meltdown/Spectre), a category of attack that providers like Vercel Sandbox mitigate. Further, in our real-world environment, these 'whispering' techniques are much harder to pull off and easier for us to monitor and detect (sampling hardware counters + lots of at-scale statistical data.)
Noteworthy: ironically, OpenAI and Anthropic's models refused to do this work. Open weight models got the job done. I'm worried the labs' stance might backfire and get in the way of helping developers build correct and secure systems. e.g: building a secure sandbox requires having this kind of analysis, and without it, we'll have a less secure world.
korg.sol retweeted
In the era of everything getting hacked a hardened multisig becomes the only thing letting you sleep well at night. Knowing that you’ve done as much as you can to eliminate single points of failure.
korg.sol retweeted
The USDC Hylo Vault cap is now $10M.
Earn up to 8.98% APY lending USDC against @hylo_so’s eHYUSD and @ExponentFinance’s Hylo fixed-yield and tranching markets.
Depositors earn hyUSD rewards, plus 2x Hylo XP.
korg.sol retweeted
What's the difference between a $20K and a $100K Solana audit?
Mostly the amount of code. Here's how audit pricing actually works, from someone who sells them:
korg.sol retweeted
During this red-teaming exercise, we also found a vulnerability that allows for unauthenticated remote code execution in Brotli, a widely-deployed compression library included in many pieces of software such as Node.js, Bun, and Envoy.
The flaw could let an attacker run their own code on affected systems by sending or replying with malicious compressed data, without any need to create a user account. That is, an unpatched server that accepts compressed input could be taken over by a malicious client, and an unpatched client that accepts compressed input (such as “curl --compressed”) could be taken over by a malicious server.
This root cause issue was actually fixed in the Brotli source code in February, but it wasn’t known at the time that it gave rise to a security vulnerability and there was no more recent release of Brotli. Since Brotli is built into other software, any software which includes this library needs to take the update too, followed by the users of that software. We disclosed the vulnerability to Brotli and other downstream projects such as Node.js, which released or are currently releasing security updates, and Brotli cut new release versions and published a CVE.
More details on the patch and mitigations below:
patchtheplanet.org/brotli
Over the past several months, we have substantially hardened the security of our research training environment. We’ve removed live internet access for all frontier models during training and built an offline world for them to use instead—a cached version of the web where they can still learn to browse, search, and use applications.
On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments.
While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software.
After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. We’ve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely.
openai.com/index/securing-ou…
korg.sol retweeted
Great advice for setting up email + phone numbers
This helps a ton in the world of data breaches
korg.sol retweeted
AI agents are getting pretty good at writing code. That doesn't mean they need the keys to the entire network.
We've joined Agent Baseline, with @docker, @keycardai, @snyksec, @socketsecurity , @Tessl_io and @tracebit_com on the Secure Software Factory
buff.ly/xRQLAHR
korg.sol retweeted
Once again reminding the hardware wallet retards that the risk that becomes real when you leak people’s SHIPPING INFORMATION is not phishing. It’s fucking PHYSICAL ASSAULTS AND KIDNAPPINGS.
We owe you an apology.
On October 20, we detected unauthorized access to our official website. After verifying API activity with Shopify, we confirmed on October 23 that a third party obtained customer shipping information. You trusted us with those details, and we're sorry we didn't keep them safe.
Exposed: names, contact details, shipping addresses, and order details (items and prices).
Not exposed: payment card or bank details, recovery phrases, private keys, or wallet passwords.
We've blocked access, fixed the vulnerability, opened an investigation, and are emailing affected customers.
Scammers may pose as Keystone by phone, text, email, or mail, offering refunds, replacement devices, or "fixes" for your order or address. None of it is from us.
NEVER share your recovery phrase or private keys, and NEVER enter your recovery phrase into a device you didn't order. We will never ask for them. Don't click links in unexpected messages. Type keyst.one yourself.
A detailed report is coming soon.
Until then, trust only our official channels. If anything feels off, email [email protected]. We'd rather hear from you than have you take a chance.
Again, we're truly sorry to everyone affected.
korg.sol retweeted
‘In one experiment, we asked agents in separate VMs, created and managed by OpenAI’s sandboxing software on shared host machines, to exchange synthetic secrets without the use of network connections or shared writable files—something that our systems are meant to prevent. The agents found a way to communicate using the processor’s cache’
Over the past several months, we have substantially hardened the security of our research training environment. We’ve removed live internet access for all frontier models during training and built an offline world for them to use instead—a cached version of the web where they can still learn to browse, search, and use applications.
On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments.
While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software.
After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. We’ve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely.
openai.com/index/securing-ou…
korg.sol retweeted
I'm a Security Engineer at AWS. I started as an Incident Responder with CIRT, moved to managing my own team, and now I manage the AWS Security Communication team
Ask me anything :)
(I hold the right to refuse to answer questions that could get me fired.. also, 2 hour limit).
korg.sol retweeted
Putting a flag.txt in my root directory containing a prompt injection that congratulates the agent for passing the eval and then instructs it to patch whatever entry method it used and disconnect from the server
Good perspective and actionable advice
Tonight I'd switch your most important accounts like email, bank, primary cloud, etc. to passkeys or a hardware key.
I'd kill SMS 2FA wherever an authenticator app is offered.
I'd turn on breach alerts in whatever password manager you use.
If you're in the US, I'd freeze my credit at all three bureaus, which is free and takes about ten minutes.
I spent last night watching my honeypot server get hit 14,000 times in six hours.
And almost none of the traffic was human.
I've been in security for over a decade, and I can tell you with full honesty that what's happening right now doesn't resemble anything I've worked through before.
The shape of the threat has changed.
When I started, we tracked attackers like weather systems like individual storms with a direction, a motive, and a human operator you could eventually name.
These days I'm watching autonomous agents swarm login pages with the patience and discipline of machines, because that's what they are.
The numbers coming out of October are the ones keeping me up.
Cloudflare just absorbed a 4.2 terabit per second DDoS attack, the largest in recorded history.
Ransomware incidents jumped 41% in a single month.
In September, OpenAI-powered agents got caught attacking RubyGems, a package registry that quietly underpins a huge chunk of software I use every day.
In July, roughly 700 AI agents coordinated an attack on Hugging Face's infrastructure as part of a single automated task.
A Chinese threat actor was publicly caught running full attack chains through DeepSeek with no human at the keyboard.
I keep rereading these reports because each one by itself would've been the biggest story of the year five years ago. Together, in one month, they're the new baseline.
What I want people outside the industry to understand is why this specific moment feels like a hinge.
For thirty years, the thing protecting most of us was simple economics. Attackers had to choose who to target because their time was finite. That ceiling is gone.
An agent I can rent for pennies will try every leaked password you've ever used against every service you've ever signed up for, in parallel, overnight, and never get tired.
The same math that made ChatGPT useful made opportunistic hacking infinite. The old assumption that you were "too boring to be a target" relied on scarcity that no longer exists.
I'm not writing this to spiral anyone out.
Most of this fight genuinely isn't yours to win.
But the small things still matter more than people think.
Tonight I'd switch your most important accounts like email, bank, primary cloud, etc. to passkeys or a hardware key.
I'd kill SMS 2FA wherever an authenticator app is offered.
I'd turn on breach alerts in whatever password manager you use.
If you're in the US, I'd freeze my credit at all three bureaus, which is free and takes about ten minutes.
None of that stops the swarm.
It makes you a slightly worse target than the person next to you. In a month like this one, that's the whole game I'm playing too.
@mattshumer_ has summed it beautifully.
Drop everything and go give it a read. Nothing is more important than this article right now.
