Matthew Green
banner
matthewdgreen.bsky.social
Matthew Green
@matthewdgreen.bsky.social
I teach cryptography at Johns Hopkins. https://blog.cryptographyengineering.com
Reposted by Matthew Green
DAVE: Open the podbay doors, ChatGPT.
CHATGPT: Certainly, Dave, the podbay doors are now open.
DAVE: The podbay doors didn't open.
CHATGPT: My apologies, Dave, you're right. I thought the podbay doors were open, but they weren't. Now they are.
DAVE: I'm still looking at a set of closed podbay doors.
June 9, 2025 at 6:04 PM
The fun thing about watching the movie 2001 in 2025 is you realize HAL is just an LLM and so *obviously* it’s going to murder its crewmembers every few flights due to malformed JSON.
January 2, 2026 at 1:27 AM
My wife has an almost mystical ability to screw up iPhones, often in ways that will persist across multiple generations of hardware. I thought she was making up the fact that her phone didn’t work (to avoid my calls) and then yesterday I watched a relatively new iPhone 15 mysteriously reboot twice.
January 1, 2026 at 7:48 PM
Reposted by Matthew Green
Happy new year.
While 2026 is an rsa modulus, it is not a product of Sophie Germain primes so it's probably a bad idea to use it.
January 1, 2026 at 8:19 AM
Where’s Waldo?
December 28, 2025 at 7:43 PM
Petition to move the winter holidays to July so we can just work through this gray time.
December 28, 2025 at 4:55 PM
I was stupid enough to buy this new AppleCare One plan for a phone I bought my daughter. Now I learn this only covers the device if it’s connected to the same Apple ID (not family plan). Have to spend Christmas unwinding this and getting a refund, what a drag.
December 25, 2025 at 6:01 PM
If we can’t solve hallucinations, OpenAI should fund a service to actually write the academic papers that ChatGPT hallucinates.
December 18, 2025 at 1:02 PM
Watching the HN folks discuss the state of user privacy in 2025 is pretty depressing. news.ycombinator.com/item?id=4630...
TikTok unlawfully tracks shopping habits and use of dating apps? | Hacker News
news.ycombinator.com
December 18, 2025 at 8:16 AM
Reposted by Matthew Green
Efficient Privacy-Preserving Blueprints for Threshold Comparison (Pratyush Ranjan Tiwari, Harry Eldridge, Matthew Green) ia.cr/2025/2253
December 18, 2025 at 1:46 AM
If you’re a cryptographer and you got one of these, send me an email.
December 18, 2025 at 1:49 AM
Imagine it’s 2013 and you see this document from the UK sent back from the future. You’ll assume something went very wrong in that timeline.
December 16, 2025 at 2:21 PM
I’m sure I should have been vibe coding with a proper IDE rather than copy/paste from an LLM, but man does AI-generated code get confusing and spaghetti after a few fixes. You have to force it to use subroutines or it’ll just produce special case after special case in one massive routine.
December 14, 2025 at 4:06 PM
Reposted by Matthew Green
1/ Yesterday’s Q2-Q3 Adversarial Threat Report by Meta was interesting in many ways. For us @citizenlab.ca, it was a blast from the past.

For the first time, Meta’s investigators attributed what in 2019 we had named Endless Mayfly - a relentless, sophisticated influence op targeting Iran’s enemies.
Burned After Reading: Endless Mayfly’s Ephemeral Disinformation Campaign - The Citizen Lab
Using Endless Mayfly as an illustration, this highlights the challenges of investigating & addressing disinformation from research & policy perspectives.
citizenlab.ca
December 12, 2025 at 4:15 PM
Reposted by Matthew Green
Europol wants to be able to break end-to-end encryption after court order (dutch article): www-security-nl.translate.goog/posting/9170...
Europol wil end-to-end encryptie na gerechtelijk bevel kunnen doorbreken
Europol wil de mogelijkheid hebben om end-to-end versleutelde communicatie van verdachten, na toestemming van de rechter, te ...
www-security-nl.translate.goog
December 12, 2025 at 8:03 PM
Reposted by Matthew Green
Last week I announced that we're finally killing off RC4 in the Windows Kerberos stack.

This has been a long time coming, so much so that we've been working on it for more than a decade, albeit off and on as we sometimes had to target other more pressing issues.

What does this mean?
Beyond RC4 for Windows authentication
As organizations face an evolving threat landscape, strengthening Windows authentication is more critical than ever.
www.microsoft.com
December 10, 2025 at 6:49 PM
So I’m on the verge of giving up and just piping my email into an LLM so I don’t have to feel guilty about not being able to read it.
December 11, 2025 at 1:28 PM
Hey, Microsoft is getting rid of RC4-based NTLM key derivation! www.microsoft.com/en-us/window...
Beyond RC4 for Windows authentication
As organizations face an evolving threat landscape, strengthening Windows authentication is more critical than ever.
www.microsoft.com
December 10, 2025 at 3:04 PM
I used the word invariant the other day in conversation and I don’t like myself for it.
December 5, 2025 at 12:50 PM
I told my son about my first programming project that other people used, a Mac Desk Accessory that could shut down the computer. So he asked ChatGPT if there was any evidence of it left online.
December 4, 2025 at 1:12 AM
If super-intelligent AI is coming (and we avoid all the bad things), I feel like philosophy is the only degree worth getting. It’s amazing to me that the tech world hasn’t figured that out.
December 1, 2025 at 3:41 PM
Trying to think of something serious to say about the “cryptographers lose the key for the cryptographer election” story and, mostly, hey: I just love that cryptographers are actually using the weird cryptography! www.nytimes.com/2025/11/21/w...
Cryptographers Held an Election. They Can’t Decrypt the Results.
www.nytimes.com
November 22, 2025 at 2:47 PM
Cryptographers Held an Election. They Can’t Decrypt the Results.
www.nytimes.com
November 22, 2025 at 2:07 AM
Reposted by Matthew Green
cloudflare's on-duty IT staff bangs on the doors which I have padlocked from the inside as I calmly break open lava lamp after lava lamp and drink the contents
November 18, 2025 at 1:59 PM