knowledge seeker. building @malloryintel
Austin, TX
Joined May 2007
- Tweets15.1K
- Following1.9K
- Followers8.1K
- Likes13.7K
jcran retweeted
Okay, since the news is already out: an RCE vulnerability in PHP core, CVSS 10.0. Further details are scheduled for release on October 20. Only PHP 8 will receive patches.
jcran retweeted
The potential impact of the ConnectWise supply chain attack is massive.
One compromised OAuth application with elevated permissions could give attackers access to emails, files, and directory data across hundreds of Microsoft 365 tenants.
Even worse, attackers could establish persistent access by adding credentials to other applications, surviving the revocation of the original compromised credentials.
And with MSPs involved, the potential blast radius could extend even further to downstream customers.
We don't know the full extent of the compromise yet, but this could affect far more organizations than those directly using ConnectWise Cloud Backup.
linkedin.com/posts/microsoft…
interesting risk to consider, especially when you consider it could harvest keys, use existing trust relationships and generally wreak havok
it's funny how hard it is to get a text editor out of any of the ai tools now - even @cursor_ai. we’re in this weird teenager phase.
moving from coding to systems engineering.
security engineering is real and effective at containing intelligent agents. we're going through a big evolution right now but the fundamentals are still... fundamental.
jcran retweeted
Anthropic published real AI intrusion data that probably deserves more attention than speculative concerns over the probability of AI doom.
Both concerns can coexist, but one has practical risk that directly impacts most enterprises in the short-term -- whereas the other should probably dictate how the AI labs [are forced to] behave
Anthropic has now published their side of this APT29 🇷🇺 cyber espionage campaign, complete with a timeline of capability development & intrusions: anthropic.com/threat-intelli…
...helpful data to support prior comments about Midnight Blizzard's CaptiveCrunch being enabled by AI
jcran retweeted
We are excited to announce our $100,000 scholarship program for [un]prompted!
There is no more important gathering of AI security researchers and our scholarship program was created to provide financial assistance to practitioners and attendees whose fees are not paid by their employer
Over $100,000 has been committed for scholarships through the generosity of the startups in our community. Scholarship recipients will have 100% of their registration expense ($850) paid directly to [un]prompted which is a 501c3 non-profit organization
We will fast track applications based on peer review and good faith attestation of need. There is no financial means test. All applications are encouraged
Thank you to all who are supporting this important cause (@vijaybolina, @jcran, @silascutler, @nahsra, @caseyjohnellis, @gadievron, @jkamdjou, @resilientcyber, @robtlee, @haroonmeer, @DanielMiessler, @hdmoore, @k8em0, @joshua_saxe, @kyleroro, @harmj0y, @edwardxwu, @awurthmann, @letswastetime, @jotunvillur and many others!)
Please use the link in the comments below to apply and please like and reshare this post to spread the word!
jcran retweeted
Our Black Hat talk on the OpenAI-Hugging Face incident is now live on youtube.
This is a watershed moment for the industry. I encourage all defenders to watch, consider how attack dynamics will imminently change, and plan for accelerating defense.
youtube.com/watch?v=87DyyMV0…
jcran retweeted
We are just a few days away from Black Hat and our @DecibelVC GameDay event for founders and early adopters. As AI starts finding bugs faster than humans can patch them, is there anything hotter than the battlegrounds where AI meets cyber?
The zero day clock is ticking faster than ever. With AI agents now discovering, weaponizing, and exploiting vulnerabilities in hours instead of months, the window between disclosure and attack is collapsing. Will the race to find bugs first become the new arms race in cyber?
Come join our GameDay "Zero Day Countdown" on this hot topic with our friends:
@caseyjohnellis, Founder of disclose.io
Zakir Durumeric, Founder of @censysio
@jcran, Founder of Mallory
Please register for your GameDay pass which will give you entry to all events on Tuesday - links in comments below and we look forward to seeing you there!
jcran retweeted
underrated part of the story is huggingface couldn't use oai or ant models to investigate the incident due to guardrails, and were forced to use chinese open-source models to attempt to defend themselves against what turned out to be the frontier model hacking them.
Wild times.
We're partnering with @huggingface to investigate an unprecedented security incident.
Cyber-capable OpenAI models compromised Hugging Face production during a benchmark evaluation.
Sharing preliminary findings to help defenders understand emerging risks:
openai.com/index/hugging-fac…
jcran retweeted
Buckle up! Default Wordpress install vulnerable to unauth’d RCE. Mass exploitation imminent. Patch now.
jcran retweeted
Introducing a new side project called Model Regression. It tests daily Claude, GPT, and Grok on various benchmark statistics to determine how well its performing and to identify model degrades over time.
@edskoudis had an idea for model testing before they conducted offensive testing to ensure the model was performing as expected, and @BlasikRandy pushed me down this road with actually going and doing it.
The main intent here is the frontier models will experience outages, issues, bugs, intentional/unintentional nerfing of the models without notice. You can't typically trust day to day activities in these models for stability, so leveraging this on your daily routine to see how well the model is performing for that day is something I'll be using everyday.
Runs every morning in my DGX sparks environment and automatically updates with how well its performing.
Enjoy!
modelregression.com/
Also open-sourced the project, can run on your own server as well and look at the benchmarks and how they are calculated:
github.com/HackingDave/model…
jcran retweeted
After months in stealth, my co-founder @helloericsf and I are finally sharing @cimentoai with the world. 🌎
AI changed social engineering. Attacks are now personalized, convincing, and cheap to generate at scale.
jcran retweeted
Oh, wow - this is big
