Eeshan retweeted
The TEE Security Handbook is a public good.
So I shouldn't be the only one to pick who works on it.
Candidates are screened & interviews start next week.
If you have built, broken, audited or deeply understand TEE systems, be a guest interviewer with me.
DMs open :)
I am hiring a chad to work on the TEE Security Handbook over at @bluethroat_labs.
It's your last chance to apply if you want to work with us.
The application is simple.
DM @iamwehi and if you get through him, you get a single interview with me and you're in :)
tee-security-handbook.blueth…
Eeshan retweeted
"We use TEEs" is not a security guarantee.
Two teams can both run TEEs and offer wildly different protection. The details are where it breaks.
4 things we check every time:
1. Physical access changes everything
If your adversary can touch the machine, we treat the TEE as compromised. Published attacks like TEE.fail are why. Whoever hosts your hardware is in your threat model.
2. Attestation doesn't mean your code is safe
A TEE will faithfully run vulnerable code. Bug in the inference engine or a dependency? Data can leak without anyone breaking the hardware isolation.
The code inside the boundary needs review too.
3. Attestation is evidence. Your verifier makes the call.
Authenticate the report. Validate the collateral. Check identity and measurements. Enforce a real policy on platform security status. Then bind that decision to the endpoint actually receiving your data.
Importing an attestation library does none of this for you.
4. Is the audited code the code that's running?
Open source is a start. You still need a verifiable link from source → build → attested measurement. Reproducible builds get you there. Without that, "audited" and "attested" can be describing two different things.
Real example from our Phala/dstack research:
CPU evidence was cryptographically verified. GPU evidence was fetched, parsed as JSON, and shown under "TEE Hardware Verified" with no GPU crypto checks.
The GPU claim had no cryptographic backing. That path has since been fixed.
bluethroatlabs.com/reveries/…
Inference providers are under attack. Securing the hardware boundary is one layer of securing the service.
More real failures and practical checks in our TEE Security Handbook:
tee-security-handbook.blueth…
nitter.gq/ilblackdragon/status/2…
Inference providers are under attack.
AI inference is now playing a critical role and is a honey pot of data and ability to inject actions back into users devices via agents for malicious actors.
NEAR AI thwarted a series of attacks on its systems and is back to normal. NEAR Protocol, Intents and other infra wasn’t affected.
TEE architecture is part of the answer to making sure that from outside even inference provider doesn’t have an access. User should verify exactly what happens with their data and prompts.
NEAR AI code is open and inspected by many partners when deciding to leverage the stack to validate exactly what runs inside TEE and ensuring privacy.
In this case there was also a vulnerability within SGLang, open source inference engine that runs inside an enclave. Even though this had a minimum disruption it still was a very
Going forward we are going increase security rigor of dependencies and also improve a “turnstile” to detect when anything out of normal runs inside TEE.
Proactive security against external attackers has became a must. We have been using Abliterated models and red teaming harness to continuously monitor and detect issues. We are also expanding scope and compute invested into this.
Eeshan retweeted
Looking forward to the same photo in Lisbon next.
We are always looking for exceptional talent to join us over at @bluethroat_labs, so we can have a bigger team XD
Eeshan retweeted
Accepted finding #5 from the @Zcash campaign :)
This was a bug where two of my most aspirational sub-agents... the "finding-escalator" & "hostile-triager" had an epic back & forth.
Their battle was one for the books. Hope to release it soon.
But it's been a helluva fun run.
Accepted finding #4 (High).
This was an OOM-kill of a critical infra piece for the @Zcash ecosystem.
Love to see maintainers find validating our report a breeze. Easy verifiability of any report is a strict standard we follow at @bluethroat_labs :)
Onwards and upwards 🚀🍻
Eeshan retweeted
Accepted finding #3.
Another High-severity issue somewhere in the @Zcash ecosystem is on the way to get patched.
Can't wait to unleash all the write-ups for all these crazy findings from our research over at @bluethroat_labs 🦅
it's bit more nuanced.
1. llms can absolutely find nasty bugs in critical software, but it's not like proving a novel theorem, it's rinse and repeat of known vulnerability patterns across huge codebases.
3. we never had the human resources to clear up the debt. in general, a few researchers had the taste, and attention to find novel bugs, think of entirely new vulnerability classes. a few hundred more could focus to sit down, search through huge code space and find bugs in stuff like v8. a few thousand could handle the lower end, the basic bugs, ~pentesters~
5. llms are pretty good at the lower and shallow middle layers, and importantly they are relentless. throw tokens huge codebases, and it will bring up all the shallow bugs, which was previously never got human attention from those few hundred researchers. have better scaffolding and infrastructure for feedback loops, it finds better findings.
5. but currently, no amount of raw tokens is going to one-shot something bugs people like manfred/xion/jann/orange(few on top of my mind). for that class of work, the operator still matters enormously. if you already understand the attack surface, you can use the llm to navigate the search space much faster. the speedup is real, but the tool is still only as good as the operator. not everyone can suddenly pwn anything, the people who already have experience move much faster.
6. this wave likely plateaus once that layer saturates.
7. but the layer is enormous, most enterprise software thinks it's secure behind defense in depth, obfuscation, and sandboxes. those won't hold and we'll see carnage before it levels off. we are going to see huge amount of cves this year.
Eeshan retweeted
Gonna say it one more time in case you did not believe me
You don't buy a coaching - you invest ... can be damn sure this amazing hacker got his money worth and then some because i do not charge this much hahaha
Eeshan retweeted
MXSS Part 2: Why Client-Side HTML Sanitization is hard
In this video, we dive into Parser Differentials, Namespace Confusion, and the Nesting Depth Limit that led to an XSS on Google and multiple DOMPurify bypasses.
youtu.be/vVwo5tW6d3w
Eeshan retweeted
During my research for the video, I compiled the MXSS Evolution and Timeline, which is now public as blog.
Awesome MXSS: github.com/msrkp/MXSS
youtube.com/watch?v=aczTceXp…
Eeshan retweeted
Hola hackers of #Hyderabad!👋
🌐 Join us for Crypto 101: An introductory session for beginners! Due to popular demand, we're switching up our usual Hacker Mixer for this special event on August 3rd.
Don't miss out! 🚀#Crypto #Blockchain #Learning
What to expect:
✨Meet cool new hacker friends
💡Learn from experts
🔎Converse, collab, & brainstorm new ideas
If you're in Hyderabad and into #cybersecurity, register today: lu.ma/mqwxkw0n
Eeshan retweeted
I did something out of my comfort zone last month. And no, it wasn't another CTF. 😃
I hosted Hyderabad's first Hacker Mixer: an informal gathering for hackers and cybersecurity enthusiasts to mingle, talk, brainstorm, and exchange ideas. 🎉
I wasn't sure what to expect, but I'm beyond happy with the turnout. 50+ people attended the Hacker Mixer. A couple of people came all the way from Vijayawada and Bangalore. Totally insane! 🤯
The official timing for the mixer was 6-8PM, but the conversations flowed until 10PM. My heart is full thinking this is only the beginning.
There's so much more to come. ✨
The most unexpected guest of the evening was 7-year-old Lakshveer. A Python enthusiast and co-founder of Power Parenting, he conducts workshops for other kids, writes blogs, does marketing, and has his own YouTube channel as well.
If the future generation has so much energy and enthusiasm, I know humanity as a whole is going in the right direction. 🌟
Cheers to more such Hacker Mixers, amazing conversations, and meaningful connections in the future. 💖
#event #hackermixer #cybersecurity
Eeshan retweeted
🔥112th edition of IWWeekly is LIVE 🔥
Read here: weekly.infosecwriteups.com/i…
📰 Featuring articles by @gurudattch, @d4d89704243, @kste_, @SchmiegSophie, @GEndignoux, @Psych0tr1a, @Krit_Sec
🧵 Trending threads by @pentest_swissky, @mcipekci, @0xTib3rius, @CupiaBart
📽️ Videos by @JohnHammond, @NahamSec, @_niteshsurana
🫂 Resource Contribution by @NikhilMemane09, @Kxddah, @thebinarybot, @tuhin1729, @Huh0x01, @_MrNiko
✒️ Newsletter formatted by @Kxddah, @bhavesharmalkar, @thebinarybot, @_EeshanV, @_MrNiko, @NandyalaVivek, @sidsec_
Eeshan retweeted
80 seats are already filled.Only 20 left. If you are a Hacker from Hyderabad and want to network with other cool hackers. Go apply now.
🌟 Exciting Announcement! 🌟
Now that the elections are over and the heat wave is gone, I am thrilled to organize Hyderabad's first exclusive hacker mixer! 🤩
Meet and network with awesome hackers, bug bounty hunters, and security professionals. Learn from their experiences, exchange ideas, and build connections to propel your career and projects forward. Students are also welcome!
I've seen such events in SF and Bangalore, but nothing in Hyd's hackerspace other than meetups. This is different - no talks, no agenda, just pure networking. Enjoy good food and great conversations.
📅 Date: 25th May
⏰ Time: 6-8 PM
📍 Venue: Book your slot here: lu.ma/hackhyd
This is my first time organizing such an event, so please help me spread the word! Limited seats only. 🚀
Please tag your infosec friends from Hyderabad.
hashtag#HackHyd hashtag#Networking hashtag#CyberSecurity hashtag#Hyderabad
Eeshan retweeted
Learn how @assetnote discovered unauthenticated XSS and RCE on Citrix products, shedding light on critical security risks.
Dive into their research in the latest issue of InfoSecWriteups Weekly Newsletter! weekly.infosecwriteups.com/i…
Eeshan retweeted
Discover how identifying unique attack surfaces can unlock critical vulnerabilities easily, as shared by @hacker_ in their latest thread 💡
Find out more in the InfoSecWriteups this weeks Weekly Newsletter! weekly.infosecwriteups.com/i…


























