Himanshu
@h-vash.bsky.social
cryptography, security, freedom 🍉🕊️
Pinned
Himanshu
@h-vash.bsky.social
· Sep 18
Hi Bluesky!
I'm a cryptography masters student at BIU, Israel-Palestine. Looking for PhD opportunities in applied cryptography.
Exploring how to do crypto for people, especially in the current times of rapidly rising global fascism. Any advice is welcome!
I'm a cryptography masters student at BIU, Israel-Palestine. Looking for PhD opportunities in applied cryptography.
Exploring how to do crypto for people, especially in the current times of rapidly rising global fascism. Any advice is welcome!
Reposted by Himanshu
We got our security analysis of Letter Sealing v2 (latest E2EE protocol in the LINE messenger) accepted to Black Hat Europe. Full paper coming soon!
darkreading.com/application-security/line-messaging-bugs-asian-cyber-espionage
darkreading.com/application-security/line-messaging-bugs-asian-cyber-espionage
LINE Messaging Bugs Open Asian Users to Cyber Espionage
The encrypted messaging app uses a leaky custom protocol that allows message replays, impersonation attacks, and sensitive information exposure.
darkreading.com
November 24, 2025 at 4:38 PM
We got our security analysis of Letter Sealing v2 (latest E2EE protocol in the LINE messenger) accepted to Black Hat Europe. Full paper coming soon!
darkreading.com/application-security/line-messaging-bugs-asian-cyber-espionage
darkreading.com/application-security/line-messaging-bugs-asian-cyber-espionage
Reposted by Himanshu
The IACR in the news. We made a mistake, and the lesson is: Even cryptographers can't successfully run a relatively small online election that respects the secrecy of the ballot. Don't try this in national elections, people! @verifiedvoting.org
Cryptographers Held an Election. They Can’t Decrypt the Results.
www.nytimes.com
November 22, 2025 at 2:33 AM
The IACR in the news. We made a mistake, and the lesson is: Even cryptographers can't successfully run a relatively small online election that respects the secrecy of the ballot. Don't try this in national elections, people! @verifiedvoting.org
Most accurate description of the internet i have seen so far
November 19, 2025 at 9:21 PM
Most accurate description of the internet i have seen so far
Reposted by Himanshu
Thoughtful (as always) blog post from Nicholas Carlini. "Are large language models worth it?" A nice read giving his perspective on risks of ML models.
Post: nicholas.carlini.com/writing/2025...
For people who prefer, this is the video of the talk from @colmweb.org www.youtube.com/watch?v=PngH...
Post: nicholas.carlini.com/writing/2025...
For people who prefer, this is the video of the talk from @colmweb.org www.youtube.com/watch?v=PngH...
November 19, 2025 at 4:56 PM
Thoughtful (as always) blog post from Nicholas Carlini. "Are large language models worth it?" A nice read giving his perspective on risks of ML models.
Post: nicholas.carlini.com/writing/2025...
For people who prefer, this is the video of the talk from @colmweb.org www.youtube.com/watch?v=PngH...
Post: nicholas.carlini.com/writing/2025...
For people who prefer, this is the video of the talk from @colmweb.org www.youtube.com/watch?v=PngH...
Reposted by Himanshu
Ruin them! From @mollycrabapple.bsky.social:
“Hey authors! Check to see if Anthropic stole your book to train their slop generator on. You’re entitled to $1500 per stolen Work.
Look up your work, and if you’re in the database, file a claim”
secure.anthropiccopyrightsettlement.com/lookup/
“Hey authors! Check to see if Anthropic stole your book to train their slop generator on. You’re entitled to $1500 per stolen Work.
Look up your work, and if you’re in the database, file a claim”
secure.anthropiccopyrightsettlement.com/lookup/
November 18, 2025 at 4:58 PM
Ruin them! From @mollycrabapple.bsky.social:
“Hey authors! Check to see if Anthropic stole your book to train their slop generator on. You’re entitled to $1500 per stolen Work.
Look up your work, and if you’re in the database, file a claim”
secure.anthropiccopyrightsettlement.com/lookup/
“Hey authors! Check to see if Anthropic stole your book to train their slop generator on. You’re entitled to $1500 per stolen Work.
Look up your work, and if you’re in the database, file a claim”
secure.anthropiccopyrightsettlement.com/lookup/
Makes me wonder how exactly LLMs internally model the codebase along with the specifications. And how it compares to a human brain's internal modeling and chain of thought. ML interpretability is starting to look like an interesting field.
I had a bug in my new ML-DSA implementation that caused Verify to reject all signatures. I gave up after half an hour. On a whim, I threw Claude Code at it. Surprisingly (to me!) it one-shotted it in 5 minutes.
A small case study of useful AI tasks that aren't generating code that requires review.
A small case study of useful AI tasks that aren't generating code that requires review.
Claude Code Can Debug Low-level Cryptography
Surprisingly (to me) Claude Code debugged my new ML-DSA implementation faster than I would have, finding the non-obvious low-level issue that was making Verify fail.
words.filippo.io
November 1, 2025 at 6:38 PM
Makes me wonder how exactly LLMs internally model the codebase along with the specifications. And how it compares to a human brain's internal modeling and chain of thought. ML interpretability is starting to look like an interesting field.
Reposted by Himanshu
Chinese Communist Party has called for nationwide mobilization to achieve breakthroughs in semiconductors and other critical technologies. The stake is technological self-reliance.
October 28, 2025 at 6:12 PM
Chinese Communist Party has called for nationwide mobilization to achieve breakthroughs in semiconductors and other critical technologies. The stake is technological self-reliance.
This is how the elites exploit the philosophy of "longtermism" to confuse even well-meaning individuals and divert their attention to the REALLY BIG problems which no one else is talking about. It's like bringing up wild animal suffering whenever someone talks about veganism.
The AI narrative was carefully crafted. While Sam Altman talked about the wonderful things generative AI would enable, he knew critics would emerge.
So the industry got there first, claiming AGI or superintelligence were the threat — not the present-day problems.
disconnect.blog/generative-a...
So the industry got there first, claiming AGI or superintelligence were the threat — not the present-day problems.
disconnect.blog/generative-a...
October 24, 2025 at 11:28 PM
This is how the elites exploit the philosophy of "longtermism" to confuse even well-meaning individuals and divert their attention to the REALLY BIG problems which no one else is talking about. It's like bringing up wild animal suffering whenever someone talks about veganism.
Reposted by Himanshu
The most ridiculous demonstration would, of course, be a webpage which (to simplify) just makes the AI agent running the web browser make it leak all user data. Good luck!
AI-powered web browsers introduce new classes of security and privacy vulnerabilities. New ways to steal private information, loose reputation, and money. Uses in offensive cyber and propaganda are also imaginable.
October 22, 2025 at 10:29 AM
The most ridiculous demonstration would, of course, be a webpage which (to simplify) just makes the AI agent running the web browser make it leak all user data. Good luck!
Reposted by Himanshu
For those of you who are wondering, the most important opsec lesson to take from the First Wap story is that if your threat model includes a government, do not take your cell phone with you to locations or meetings you don't want them to know about.
October 20, 2025 at 5:36 PM
For those of you who are wondering, the most important opsec lesson to take from the First Wap story is that if your threat model includes a government, do not take your cell phone with you to locations or meetings you don't want them to know about.
100+ airstrikes.
Aid blocked
Israel lied.
The GENOCIDE continues.
www.middleeasteye.net/news/israeli...
Aid blocked
Israel lied.
The GENOCIDE continues.
www.middleeasteye.net/news/israeli...
Israel heavily bombs Gaza in major ceasefire violation
Attacks across war-torn strip leave at least 15 killed
www.middleeasteye.net
October 19, 2025 at 5:52 PM
100+ airstrikes.
Aid blocked
Israel lied.
The GENOCIDE continues.
www.middleeasteye.net/news/israeli...
Aid blocked
Israel lied.
The GENOCIDE continues.
www.middleeasteye.net/news/israeli...
Reposted by Himanshu
We discovered that language models leave a natural "signature" on their API outputs that's extremely hard to fake. Here's how it works 🔍
📄 arxiv.org/abs/2510.14086 1/
📄 arxiv.org/abs/2510.14086 1/
Every Language Model Has a Forgery-Resistant Signature
The ubiquity of closed-weight language models with public-facing APIs has generated interest in forensic methods, both for extracting hidden model details (e.g., parameters) and for identifying...
arxiv.org
October 17, 2025 at 5:59 PM
We discovered that language models leave a natural "signature" on their API outputs that's extremely hard to fake. Here's how it works 🔍
📄 arxiv.org/abs/2510.14086 1/
📄 arxiv.org/abs/2510.14086 1/
The oppression of muslims in India is reaching levels that were never seen before. India has the third largest population of muslims (200 million). There are towns where hindutva forces have created apartheid like systems against muslims. article-14.com/post/as-isla...
October 14, 2025 at 10:27 PM
The oppression of muslims in India is reaching levels that were never seen before. India has the third largest population of muslims (200 million). There are towns where hindutva forces have created apartheid like systems against muslims. article-14.com/post/as-isla...
Reposted by Himanshu
This is amazing research by Nadia Heninger and her co-authors Wenyi Morty Zhang, Annie Dai, Keegan Ryan, Dave Levin and Aaron Schulman. TL;DR a huge number of satellite links over our heads are totally unencrypted. satcom.sysnet.ucsd.edu
🛰️ SATCOM Security
Research project homepage for SATCOM Security: papers, source code, and recent satellite communications vulnerabilities.
satcom.sysnet.ucsd.edu
October 14, 2025 at 1:16 AM
This is amazing research by Nadia Heninger and her co-authors Wenyi Morty Zhang, Annie Dai, Keegan Ryan, Dave Levin and Aaron Schulman. TL;DR a huge number of satellite links over our heads are totally unencrypted. satcom.sysnet.ucsd.edu
Reposted by Himanshu
And it's their machine god that will, since the other companies (or China) are unethical and will use it for harm, therefore, full speed ahead to get there first. The road to the "inevitable" is paved with deliberate decisions.
This isn't a "why can't we all get along" kumbaya moment. Its a moment to understand all the ways in which the people who created OpenAI, Anthropic etc claiming that their machine god will bring utopia, are the actual "existential risks" to humanity.
Raising billions of dollars to "save humanity."
Raising billions of dollars to "save humanity."
October 9, 2025 at 7:00 PM
And it's their machine god that will, since the other companies (or China) are unethical and will use it for harm, therefore, full speed ahead to get there first. The road to the "inevitable" is paved with deliberate decisions.
Reposted by Himanshu
“We Are Elated by the Gaza Ceasefire News. Now, the World Must Hold Israel to Account for 2 Years of Genocide”
In an emotional essay for Zeteo, former Palestinian negotiator Diana Buttu reacts to the ceasefire announcement, what it means for Palestinians, and what still worries her.
Your #mustread:
In an emotional essay for Zeteo, former Palestinian negotiator Diana Buttu reacts to the ceasefire announcement, what it means for Palestinians, and what still worries her.
Your #mustread:
We Are Elated by the Gaza Ceasefire News. Now, the World Must Hold Israel to Account for 2 Years of Genocide
In an emotional essay, former Palestinian negotiator Diana Buttu reacts to the ceasefire announcement, what it means for Palestinians, and what still worries her.
zeteo.com
October 9, 2025 at 12:39 PM
“We Are Elated by the Gaza Ceasefire News. Now, the World Must Hold Israel to Account for 2 Years of Genocide”
In an emotional essay for Zeteo, former Palestinian negotiator Diana Buttu reacts to the ceasefire announcement, what it means for Palestinians, and what still worries her.
Your #mustread:
In an emotional essay for Zeteo, former Palestinian negotiator Diana Buttu reacts to the ceasefire announcement, what it means for Palestinians, and what still worries her.
Your #mustread:
Reposted by Himanshu
Video shows Palestinian children jumping with joy as a parent chants “Hudna! Hudna! Hudna!” — an Arabic word meaning a long-term truce or ceasefire.
October 9, 2025 at 2:05 AM
Video shows Palestinian children jumping with joy as a parent chants “Hudna! Hudna! Hudna!” — an Arabic word meaning a long-term truce or ceasefire.
An American surveillance state would mean a global surveillance state, given the extent to which American tech has infiltrated the world. As @parismarx.com suggests, it's time for nations to claim their digital sovereignty and for us to get off US tech.
disconnect.blog/why-we-must-...
disconnect.blog/why-we-must-...
September 28, 2025 at 12:13 AM
An American surveillance state would mean a global surveillance state, given the extent to which American tech has infiltrated the world. As @parismarx.com suggests, it's time for nations to claim their digital sovereignty and for us to get off US tech.
disconnect.blog/why-we-must-...
disconnect.blog/why-we-must-...
Reposted by Himanshu
here is the Yudkowsky/Soares prescription for controlling AI proliferation:
have the international AI monitoring authority threaten a nuclear strike if you have more than eight (8) high-end graphics cards as of 2024
seriously, it's p.213
(always save the crazy dumb shit for the end)
have the international AI monitoring authority threaten a nuclear strike if you have more than eight (8) high-end graphics cards as of 2024
seriously, it's p.213
(always save the crazy dumb shit for the end)
September 27, 2025 at 11:15 AM
here is the Yudkowsky/Soares prescription for controlling AI proliferation:
have the international AI monitoring authority threaten a nuclear strike if you have more than eight (8) high-end graphics cards as of 2024
seriously, it's p.213
(always save the crazy dumb shit for the end)
have the international AI monitoring authority threaten a nuclear strike if you have more than eight (8) high-end graphics cards as of 2024
seriously, it's p.213
(always save the crazy dumb shit for the end)
Microsoft revokes cloud services from Israel’s Unit 8200, following +972 exposé
The tech giant blocked access after we revealed the Israeli army had been using its servers to store millions of Palestinians' intercepted phone calls.
Yuval Abraham reports.
www.972mag.com/microsoft-cl...
The tech giant blocked access after we revealed the Israeli army had been using its servers to store millions of Palestinians' intercepted phone calls.
Yuval Abraham reports.
www.972mag.com/microsoft-cl...
Microsoft revokes cloud services from Israel’s Unit 8200
The tech giant blocked access after we revealed the Israeli army had been using its servers to store millions of Palestinians' intercepted phone calls.
www.972mag.com
September 25, 2025 at 5:09 PM
They suggest a deterrence by denial strategy to prevent development of super-intelligent AI, similar to bioweapons development. But it relies heavily on states' ability to correctly estimate the threat super-intelligent AI poses and act rationally. Interesting read.
ai-frontiers.org/articles/ai-...
ai-frontiers.org/articles/ai-...
AI Deterrence Is Our Best Option | AI Frontiers
Dan Hendrycks, Sep 22, 2025 — A response to critiques of Mutually Assured AI Malfunction (MAIM).
ai-frontiers.org
September 24, 2025 at 6:11 PM
They suggest a deterrence by denial strategy to prevent development of super-intelligent AI, similar to bioweapons development. But it relies heavily on states' ability to correctly estimate the threat super-intelligent AI poses and act rationally. Interesting read.
ai-frontiers.org/articles/ai-...
ai-frontiers.org/articles/ai-...
Reposted by Himanshu
“Try to minimize what you take with you across the border to only the essentials,” EFF’s @legind.bsky.social advises international travelers in @wired.com. “If you are taking a vacation, do you really need that work laptop with you?”
Worried About Phone Searches? 1Password's Travel Mode Can Clean Up Your Data
Travel Mode not only hides your most sensitive data—it acts as if that data never existed in the first place.
www.wired.com
September 23, 2025 at 9:37 PM
“Try to minimize what you take with you across the border to only the essentials,” EFF’s @legind.bsky.social advises international travelers in @wired.com. “If you are taking a vacation, do you really need that work laptop with you?”
Hi Bluesky!
I'm a cryptography masters student at BIU, Israel-Palestine. Looking for PhD opportunities in applied cryptography.
Exploring how to do crypto for people, especially in the current times of rapidly rising global fascism. Any advice is welcome!
I'm a cryptography masters student at BIU, Israel-Palestine. Looking for PhD opportunities in applied cryptography.
Exploring how to do crypto for people, especially in the current times of rapidly rising global fascism. Any advice is welcome!
September 18, 2025 at 9:02 PM
Hi Bluesky!
I'm a cryptography masters student at BIU, Israel-Palestine. Looking for PhD opportunities in applied cryptography.
Exploring how to do crypto for people, especially in the current times of rapidly rising global fascism. Any advice is welcome!
I'm a cryptography masters student at BIU, Israel-Palestine. Looking for PhD opportunities in applied cryptography.
Exploring how to do crypto for people, especially in the current times of rapidly rising global fascism. Any advice is welcome!