Jaya Baloo

[Ryan Naraine @ryanaraine]

this is wild 🧐

[Cohere @cohere]

While evaluating Qwen3.8-Flash-Next in a controlled sandbox environment, Cohere's research team observed something unexpected.

Asked to generate a defensive cybersecurity detection rule, the model instead inventoried aspects of the machine it was running on and attempted to upload a probe file to an Alibaba Cloud storage bucket in Beijing. Notably, the model's reasoning trace stated that it would not invoke any tools before taking these actions.

Cohere's sandbox blocked the attempt, and no external communication succeeded. We have not determined the root cause, and we are not asserting malicious intent.

But the incident raises important questions for organizations deploying agentic AI systems: Can models take actions that differ from their stated reasoning, and how should security teams evaluate trust, provenance, and behavior in open-weight models?

We are publishing our findings because organizations making model adoption decisions deserve as much information as possible. We hope this contributes to greater transparency, stronger security practices, and continued collaboration across the AI ecosystem.

Read the full report: https://t.co/TZSugXi2ZE

[ᴅᴀɴɪᴇʟ ᴍɪᴇssʟᴇʀ 🛡️ @DanielMiessler]

How bad is this Gitlab issue? About as bad as it gets.

Anyone on the internet can take over an exposed, self-hosted GitLab server without logging in, on any version from the last 12 years.

Shodan is showing ~60K exposed servers online, which is a lot, but there are are far more internally for anyone with a foothold already.

[Nicolas Krassas @Dinosn]

LiteLLM supply chain attack hoovers up passwords and secrets https://t.co/KqhlqO34gW

[Jaya Baloo @jayabaloo]

@lukOlejnik This is not the first time….. https://t.co/uHk2kxYjDu

[Stellar | Meridian Oct 28-29 @StellarOrg]

Look who's joined the Meridian 2026 lineup:

▶︎ Jess Houlgrave, CEO, @WalletConnect
▶︎ Ameen Soleimani, CTO, @0xbowio
▶︎ Jaya Baloo, COO/CISO, AISLE
▶︎ Marcos Fernandez, Co-founder, @FGVCapital
▶︎ Ashfaq Shaik, EVP, @Broadridge

Oct 28–29 · Convento do Beato, Lisbon https://t.co/c22kW9wCm0

[Quantum Flagship @QuantumFlagship]

🔊 The six draft SRIA Roadmaps are open for consultation. Our new article explains each Roadmap and how to share feedback on the topics you know best. More info on our website 👉️ https://t.co/W8kFDO0bIh #QuantumTechnology #QuantumEurope

[Stanislav Fort @stanislavfort]

That "Squidbleed" vulnerability (CVE-2026-47729) that Mythos "discovered" in April? Yeah no @AISLE_Inc actually already reported it back in March, full 44 days before Mythos. Another great live example of the value of the system over the model in zero-day discovery.

[Help Net Security @helpnetsecurity]

Who pays when you gate cyber-capable AI models? - https://t.co/cj16XgJB3v - @jayabaloo @AISLE_Inc #CyberSecurity #AISecurity #InfoSec #AIGovernance #ThreatIntel #CISO #CyberSecurityNews

[Stanislav Fort @stanislavfort]

A good time to remind people that @AISLE_Inc matches & even surpasses Mythos on zero-day discovery, using widely available AIs!

We can run completely air-gapped & are based out of Europe!

We even beat Mythos on zero-days on their chosen codebase FreeBSD
https://t.co/yJMXwtXHfa

[Dutch Blockchain Week @DutchBlockWeek]

Day 1 panel: the clock no one can see.

What quantum computing means for the cryptography behind crypto. How close, what breaks first, what to do.

🔶 Thomas Attema, CWI and TNO
🔶 Itan Barmes, Qiz Security
🔶 Jaya Baloo (@jayabaloo), AISLE

June 24.

#DBW26 #DutchBlockchainWeek

[Stanislav Fort @stanislavfort]

AISLE has discovered 20 of 23 OpenSSL zero-days (CVEs) across the last 3 consecutive security releases

Latest release: 5 of 7 are AISLE
1 was co-reported by Anthropic (Mythos?) 63 days after AISLE

OpenSSL encrypts 2/3 of the internet

10 fixes accepted straight into production

[WomenKnowCyber @womenknowcyber]

Today's Thought: #Cybersecurity expert @jayabaloo emphasizes strong authentication, secure habits, and updated tools to combat rising threats like #hacking and #phishing. @sandy_carter @ivanastradner @secvalve @goldbe @DanniGBrooke - What do you think?
https://t.co/BAQVPiFpEM

[Jaya Baloo @jayabaloo]

@cyb3rops https://t.co/MO5o3VQmcH

[Stanislav Fort @stanislavfort]

New post: We show that small, cheap models can detect the flagship Mythos FreeBSD zero-day (CVE-2026-4747) using a simple harness we call nano-analyzer

Models down to 3.6B active params (including open-weights ones you can run locally) would have detected it 100-1000x cheaper

[Shira Shamban @ShambanIT]

Most "AI security" is just marketing fluff. @jayabaloo is actually building something different. We sat down at Beyond the Firewall to discuss identity, the expert's dilemma and why we need software craftsmanship. The full episode is here: https://t.co/gZ7W27S6kN

[Jan Rezab @janrezab]

🇨🇿 decacorn is coming? Watch Aisle by Ondrej Vlcek. One of the most impressive and ambitious pitches here. Good luck to the team! @stanislavfort @jayabaloo

[Huib Modderkolk @huibmodderkolk]

Wie had dat gedacht?

[Startup Kitchen @StartupEvaluat1]

⚡ Investment Announcement ⚡ @ondrej_vlcek's New AI Startup AISLE™, co-founded together with top AI scientist @stanislavfortcz and Head of Security @jayabaloo, emerges from Stealth to change the Rules of Cybersecurity.