QM

514 posts

QM banner
QM

QM

@qasimmith

CEO of @depthfirstlabs. On a mission to secure the world's software. ex early databricks, AWS.

San Francisco, CA Katılım Haziran 2024
495 Takip Edilen1.8K Takipçiler
QM retweetledi
Dino A. Dai Zovi
Dino A. Dai Zovi@dinodaizovi·
Because regex-triggered vulnerabilities depend on the specific regex input, they are especially difficult for static analyzers (and humans) to find. This is impressive.
Zhenpeng (Leo) Lin@Markak_

NGINX rift: We autonomously discovered this 18 yr old heap overflow (CVE-2026-42945) in @nginx impacting version 0.6.27 to 1.30.0. If you use rewrite and set directive, you maybe impacted! Please update your NGINX or change the config to mitigate it. Read more at depthfirst.com/nginx-rift

English
0
7
26
5.8K
QM retweetledi
Het Mehta
Het Mehta@hetmehtaa·
Can’t patch Nginx right now? Here’s your stopgap. CVE-2026-42945 triggers through unnamed regex captures in rewrite rules. Change this:  rewrite ^/user/([0-9]+)$ /profile?id=$1;  To this:  rewrite ^/user/(?<id>[0-9]+)$ /profile?id=$id;  Not a fix. A workaround.
Het Mehta@hetmehtaa

Every 3rd website you visit runs Nginx. 18,959,833 of them can be hijacked right now. A bug from 2008 just got a working exploit. CVE-2026-42945 (CVSS 9.2) No login. No access. Just one HTTP request. → Heap overflow → Worker process → RCE Patch ASAP to Nginx 1.31.0 or 1.30.1 PoC is already out: github.com/DepthFirstDisc…

English
2
19
121
17.9K
QM
QM@qasimmith·
@kkrypt0nn @IntCyberDigest you can actually exploit this with ASLR on - it will take longer but with AI very possible and not difficult. There are probably close to a million servers out there with ASLR disabled but 10's+ million nginx with ASLR. I'm sure an attacker would spend the extra $x in compute.
English
0
0
0
60
Krypton
Krypton@kkrypt0nn·
@IntCyberDigest Missing an important part which is that it affects installations that have ASLR disabled. The PoC specifically disables it.
English
2
0
36
16.4K
International Cyber Digest
International Cyber Digest@IntCyberDigest·
‼️🚨 MAJOR IMPACT: AI just found an 18-year-old NGINX critical remote code execution vulnerability. It has been disclosed on GitHub including PoC code. - Affects NGINX 0.6.27 through 1.30.0 - Triggered via the rewrite and set directives in config - Update NGINX ASAP - NGINX is a widely used HTTP web server, be sure to check its prevalence in other products
International Cyber Digest tweet media
English
86
402
2.6K
931.5K
QM
QM@qasimmith·
@IntCyberDigest @SergioOSINT you can actually exploit this with ASLR on - it will take longer but with AI very possible and not difficult..
English
0
0
1
61
QM
QM@qasimmith·
@vxunderground "Theoretically, we could leverage this design to leak ASLR by progressively overwriting pointers byte by byte. In this post, we discuss the exploitation technique assuming ASLR has already been bypassed." It will take time but you can theoretically do this with AI.
English
1
0
6
2.3K
vx-underground
vx-underground@vxunderground·
> wake up > take a shit > get out of bed > check internet > nginx rce > look inside > requires ASLR disabled
vx-underground tweet media
English
53
137
3.1K
101.1K
QM
QM@qasimmith·
@Polymarket This doesn't make sense.
English
0
0
1
40
Polymarket
Polymarket@Polymarket·
JUST IN: OpenAI announces support for the creation of a "global governance body" for artificial intelligence led by the U.S. & China.
English
207
282
2.2K
265.8K
QM retweetledi
Zhenpeng (Leo) Lin
Zhenpeng (Leo) Lin@Markak_·
Using the same system, we found NGINX RCE, Linux LPE, Chrome RCE, FFmpeg RCE and a lot of other critical Vulnerabilities, feel free to try it out! We are trying our best to help secure OSS!
QM@qasimmith

Today we're launching the Open Defense Initiative: up to $5 million in @depthfirstlabs credits for critical open source projects to find and fix real, exploitable vulnerabilities. The timing matters: frontier models can autonomously discover and exploit vulnerabilities in widely-reviewed codebases. Open source models will catch up soon, and when they do, bad actors will have unfiltered access to these capabilities. We have a narrow window to harden critical software before that happens. This is the time to act, but until today frontier-level security, like what Mythos offers, has been reserved for a handful of large companies who are required to pay a lot for access. depthfirst is not only comparable in performance but also goes significantly beyond surface level findings, highlighting real, exploitable vulnerabilities due to its understanding of the system’s context and ability to verify like an attacker would. depthfirst found vulnerabilities in FFmpeg that Mythos missed, at a tenth of Anthropic's self reported spend. We want every defender to have these capabilities, starting with the open source projects the world runs on. If you maintain a critical open source project, apply for Open Defense credits through the form in the comments.

English
0
11
72
11.8K
QM
QM@qasimmith·
This is why we've launched the Open Defense Initiative: we want every critical open source project to proactively find and fix vulnerabilities with @depthfirstlabs, for free. You don't need a frontier model or an enterprise budget to have access to frontier-level security. OSS can apply through the link in the comments.
Zhenpeng (Leo) Lin@Markak_

NGINX rift: We autonomously discovered this 18 yr old heap overflow (CVE-2026-42945) in @nginx impacting version 0.6.27 to 1.30.0. If you use rewrite and set directive, you maybe impacted! Please update your NGINX or change the config to mitigate it. Read more at depthfirst.com/nginx-rift

English
1
0
11
399
QM
QM@qasimmith·
More results of our commitment to secure open source software. @depthfirstlabs autonomously found, validated, and generated a fix for a critical heap overflow in NGINX’s rewrite module leading to a Remote Code Execution. NGINX powers a large portion of global web traffic, and is used by major companies to run and secure their web services. This code had been there for 18 years and run countless times before we found the vulnerability. This is why the Open Defense Initiative is so important. Soon anyone will have access to open-weights models that make vulnerability detection much easier and scalable. We have a short window of time to help secure open source projects before then, and we’ve committed $5m to do so. Thank you to the NGINX maintainers for collaborating with depthfirst and patching this vulnerability so promptly. Find links for NGINX Rift, the Open Defense Initiative, and Forbes coverage on the initiative in the comments 👇
English
2
2
9
295
QM
QM@qasimmith·
@depthfirstlabs If you use NGINX (most web services out there do), please update your version immediately.
English
0
0
0
42
depthfirst
depthfirst@depthfirstlabs·
depthfirst autonomously discovered, verified, and generated a patch for NGINX rift, an 18 year old heap overflow (CVSS 9.2). It leads to an RCE and is affecting most of the global web traffic. Follow the link in the comments to learn more.
Zhenpeng (Leo) Lin@Markak_

NGINX rift: We autonomously discovered this 18 yr old heap overflow (CVE-2026-42945) in @nginx impacting version 0.6.27 to 1.30.0. If you use rewrite and set directive, you maybe impacted! Please update your NGINX or change the config to mitigate it. Read more at depthfirst.com/nginx-rift

English
2
2
7
2.6K
Alex Shan
Alex Shan@alexshander03·
We’re launching @JudgmentLabs today and announcing $32M in funding. As AI agents take on more of the work that creates economic value, they generate massive amounts of production data: the clearest record of how they behave with users, software, and the real world. Judgment builds infrastructure for improving AI agents from production data.
English
208
156
1K
3.5M
Kirsten Green
Kirsten Green@kirstenagreen·
"Depthfirst's AI found major flaws in tools that help run much of the internet—all for a tenth of the cost of Anthropic’s comparable model Mythos." Today, @depthfirstlabs is launching Open Defense Initiative, a program offering companies and open source developers a total of $5M in credits to use its AI & find vulnerabilities in their code: depthfirst.com/open-defense
QM@qasimmith

Today we're launching the Open Defense Initiative: up to $5 million in @depthfirstlabs credits for critical open source projects to find and fix real, exploitable vulnerabilities. The timing matters: frontier models can autonomously discover and exploit vulnerabilities in widely-reviewed codebases. Open source models will catch up soon, and when they do, bad actors will have unfiltered access to these capabilities. We have a narrow window to harden critical software before that happens. This is the time to act, but until today frontier-level security, like what Mythos offers, has been reserved for a handful of large companies who are required to pay a lot for access. depthfirst is not only comparable in performance but also goes significantly beyond surface level findings, highlighting real, exploitable vulnerabilities due to its understanding of the system’s context and ability to verify like an attacker would. depthfirst found vulnerabilities in FFmpeg that Mythos missed, at a tenth of Anthropic's self reported spend. We want every defender to have these capabilities, starting with the open source projects the world runs on. If you maintain a critical open source project, apply for Open Defense credits through the form in the comments.

English
3
0
15
3.3K