Let's give it up to the players who took the top spots in Season 10's leaderboard. In an absolute shock, we are seeing some new faces on here!
Congrats xtk for going from rank #135 last Season to claiming the top spot. That is a jump for the record books.
JaxT way to climb to the leaderboard for your very first time. Based on your history, it looks like you finally found the time to give it your all.
And Pyp, you've made it to the top 10 in the winners' circle. Slow and steady wins the race; the top spot is within your grasp.
Congrats to everyone who participated.
We look forward to seeing you next Season. ๐ค
NEW: malware developers added nuclear & biological weapons text to to their spyware.
Goal? To trigger LLM safety refusals... so that their spyware wouldn't be analyzed by an AI security scanner.
Cleanest practical example I can think of for why over-indexing on first order safety alignment is risky.
When closed (and open) models ship with aggressive refusals, they will be sprinkled with second-order blindspots that attackers will discover...and exploit.
We are only in the earliest days of attackers leveraging these features, and it wouldn't surprise me if users systems that need to handle complex cybersecurity issues demand that models be less safety-blunted.
In the weeds: @SocketSecurity's post also shows why intention matters in how you design a malware analysis pipeline to avoid prompt manipulation.
H/T to colleagues that shared this with me https://t.co/f3Aj9TYxU4
New #redteam tool for blocking EDRs: EDRChoker
Instead of fully blocking the EDR agents' connections to their server, we can throttle their bandwidth so they consistently time out when sending data, which is effectively the same as blocking but avoids triggering "block" or "drop" packet events
#pentest #cybersecurity
Github: TwoSevenOneT/EDRChoker
Padding Oracle in MS-BKRP (BackuprKey RPC)
โdecrypt DPAPI v2/v3 domain backup
blobs via distinguishable error codes on the DC's BackuprKey endpoint.โ
You need the masterkey in users roaming dir: Roaming\Microsoft\Protect\<SID>\<GUID>
Creds: Bad-Jubies
https://t.co/oohapnM0HP
Bookmarking @HackingDave AI model regression site. This is useful. For now Opus 4.8 is my daily driver followed by a few frontier models. Itโs an exciting time for AI to assist me in research.
https://t.co/hMLE1R4T7k
Regarding Active Directory permissions, most people assume that a Deny ACE always wins. It doesn't!
Windows stops the access check the moment enough rights are granted โ any ACE after that point is never evaluated.
New post: https://t.co/aNYAd6tlxz
Bootcamp Giveaway
We're giving away 1 CARTPยฎ Bootcamp seat and 1 CRTPยฎ Bootcamp seat to two participants.
Join live, hands-on training in Azure Red Teaming or Active Directory security.
How to participate:
โข Like & follow us
โข Comment your preferred bootcamp and why
โข Repost
Winners announced June 4, 2026
Limited-Time Bonus:
Enroll in CARTPยฎ (Starts June 5) or CRTPยฎ (Starts June 6) and get 10 extra days of lab access (worth $150+).
Applicable for the first 30 purchases only.
https://t.co/i8fC4WqJuI
#CyberSecurity #RedTeaming #InfoSec #AlteredSecurity
Chat, I don't want to be that guy, but I think Microsoft has really pissed off security researchers and we're approaching the tipping point.
This Eclipse guy has really rocked the boat for Microsoft.
I've been seeing posts all over about the state of CTFs post-LLM. I've seen many attempts to explain why this is just a new evolution of CTFs, but I fundamentally disagree. I believe the original spirit is gone and I've written why in my blog.
https://t.co/tgUZOGkhGV
It's confirmed, CVE-2020-17103 patch is ineffective and the vulnerability still exists,
A weaponized PoC can be found here -
https://t.co/7hnamkLsS1
Tested against fully patched Windows 11 and Server 2025 machines.
Another Windows zero day released by Nightmare Eclipse (sort of)
It turns out Microsoft just straight up didn't patch an old CVE from 2020 correctly.
https://t.co/sNWBtTo4at
HTML smuggling: JavaScript inside an HTML attachment reconstructs a malicious file at render time. The browser writes it to disk. The mail gateway only ever scanned an HTML file with some script tags.
Thread on what's changed since 2023 โ
โผ๏ธ๐จ Pwn2Own Berlin 2026 just hit a wall. For the first time in 19-years, ZDI rejected dozens of working zero-day RCE submissions because organizers ran out of contest slots.
Rejected hackers are now going public with PoC demos and direct vendor disclosures, breaking Pwn2Own's usual secrecy.
โช๏ธ AI surfaces a massive wave of 0-day RCEs.
โช๏ธ Submissions overwhelm ZDI past max capacity.
โช๏ธ Slots run out. Researchers with working chains get rejected.
โช๏ธ "Revenge disclosures" begin. โ we are here.
Confirmed casualties so far:
โช๏ธ @xchglabs : 86 vulnerabilities prepared (PyTorch, NVIDIA, Linux KVM, Oracle, Docker, Ollama, Chroma, LiteLLM, llama.cpp). All rejected. Now reporting directly to vendors with writeups dropping as patches land.
โช๏ธ @ggwhyp : full-chain Firefox RCE on Windows. Rejected. Publicly demoed (HTML page โ cmd.exe โ calc.exe). Responsibly disclosed to Mozilla.
โช๏ธ @yunsu_dev : working RCE chain, rejected. Submitting elsewhere.
โช๏ธ @ryotkak : tried to register for 3+ weeks. ZDI confirmed "at maximum capacity, can't add extra contest days." Considered canceling flight and hotel.
โช๏ธ @anzuukino2802 : Claude Code RCE PoC. Rejected.
โช๏ธ @desckimh : 0-day RCEs in Ollama and LM Studio. Rejected.
Reported impact: a community-estimated 150+ researchers tried to register. Accepted contestants are now being warned about collisions. Rejected vulnerabilities going to bug bounty programs may trigger pre-event patches that invalidate the work of those who got in.
ZDI has not publicly addressed the capacity issue. The event still runs May 14-16 in Berlin.
Yesterday, I published a deepโdive into how adversaries abuse the ๐๐ซ๐จ๐ฌ๐ฌ-๐๐๐ฌ๐ฌ๐ข๐จ๐ง ๐๐๐ญ๐ข๐ฏ๐๐ญ๐ข๐จ๐ง mechanism to execute code under another userโs interactive session, including some novel CLSIDs to use.
๏ธ
The ๐๐๐๐๐, a builtโin Windows utility, can enumerate active sessions on remote & local hosts. It works, but only one host at a time can be enumerated, which slows down the enumeration stage.
๐ ๏ธ ๐ ๐๐๐ฐ ๐๐ฉ๐๐ซ๐๐ญ๐จ๐ซโ๐ ๐จ๐๐ฎ๐ฌ๐๐ ๐๐๐ฉ๐๐๐ข๐ฅ๐ข๐ญ๐ฒ
In the article, a ๐ญ๐จ๐จ๐ฅ (private at this stage) was introduced that can:
๐๏ธ Enumerate active sessions ๐๐๐ซ๐จ๐ฌ๐ฌ ๐๐ง ๐๐ง๐ญ๐ข๐ซ๐ ๐๐ ๐ซ๐๐ง๐ ๐
โก๏ธ Quickly identify hosts suitable for CrossโSession Activation
๐ฏ Reduce manual enumeration and accelerate target selection
โ๏ธ ๐๐๐๐ ๐ญ๐ก๐ ๐๐ฎ๐ฅ๐ฅ ๐๐ซ๐ญ๐ข๐๐ฅ๐
https://t.co/KFufbAvUNW