@neilmadden@infosec.exchange

Neil Madden @neilmadden@infosec.exchange · 3d ago

This is a really interesting and thorough experience report of deploying Device-Bound Session Credentials. My feeling is that DBSC needs another spec iteration or two before being ready for wide deployment. My guess is that if it does catch on, we’ll see attacks move from exfiltrating session cookies to proxying requests through the user’s browser/device - as we’ve seen in the past eg with HttpOnly cookies. Without an equivalent of WebAuthn’s UserPresence bit, I’m not sure this is a long-term solution. See https://neilmadden.blog/2021/03/20/towards-a-standard-for-bearer-token-urls/ for my thoughts on an alternative approach. https://scotthelme.co.uk/everything-i-learned-shipping-device-bound-session-credentials/

0

Neil Madden @neilmadden@infosec.exchange · Aug 07, 2026

Ha! A friend of mine is using ChatGPT to write a personal statement for a job application (Yes, I am judging them). They sent me a screenshot of it telling them “I actually don’t want to write this statement tonight.” Literally the deadline for applications… AI will multiply productivity, except when it’s being too sassy, apparently. Wat.

0

Neil Madden @neilmadden@infosec.exchange · Aug 03, 2026

Discovered this beautiful Big Star cover of Don’t Worry Baby earlier. (From a comment on an Instagram reel of someone else doing a pretty good cover too). https://youtu.be/PqlHiWrVGag

0

Neil Madden @neilmadden@infosec.exchange · Aug 03, 2026

RE: https://fosstodon.org/@bsdphk/117032189389327188 “NVD quickly flagged these as critical, and CISA's ADP agreed. But when JFrog security researchers dug in to verify, the claims fell apart.” More garbage from NVD. Do they spend more than 5 minutes “enriching” CVEs?

0

1

0

Neil Madden @neilmadden@infosec.exchange · Aug 03, 2026

Replying to @dkf@cyberplace.social

@dkf@cyberplace.social I worked on OpenAM at ForgeRock for many years, which was a fork of Sun’s OpenSSO - which dated back to a Sun/Netscape collab from the late 90s (iPlanet). At some point someone unearthed an old Sun manual for a very early version, and I think the office was filled with the sound of pennies dropping for weeks afterwards…

0

Neil Madden @neilmadden@infosec.exchange · Aug 02, 2026

Many times in my career I’ve found myself working on software that is 15-20+ years old and has seen multiple generations of developers come and go. What strikes me is that the VCS history is never complete. Large swathes of early design decisions and context are missing. I find there’s often a sort of folklore history that then develops, with collective imagining of why certain features exist or are designed the way they are. Often these post-hoc rationales don’t really hold up to scrutiny, or are (IMO) anachronistic. Hopefully these days we are a bit more aware of the importance of preserving history when switching version control software. But I do still wonder how much historical context gets lost in forgotten wikis, bug trackers etc. This is one of the reasons I like fossil - it’s all part of the version history. ADRs in the repo are also a good idea, but in my experience lots of architectural decisions happen at the org-level rather than individual project level, so again end up on a wiki instead. Anyone seen examples of multi-decade projects that have successfully maintained accurate knowledge of early design decisions?

1

2

0

Neil Madden @neilmadden@infosec.exchange · Jul 31, 2026

ICYMI: The bugs Anthropic #mythos found in crypto code are interesting, but not astonishing. Write ups from cryptographers: Matthew Green: https://blog.cryptographyengineering.com/2026/07/29/some-notes-about-anthropics-new-results/ JP Aumasson: https://www.bfswa.blog/p/llms-wont-break-symmetric-crypto tl;dr - the break of HAWK post-quantum signature scheme is not fatal but makes it much less attractive as a standard once parameters are adjusted. The “attack” on AES is of theoretical interest only. But humans missed them, so there is real value here. #cryptography #ai

A Few Thoughts on Cryptographic Engineering

Some thoughts about Anthropic’s new cryptanalysis results

Yesterday Anthropic published two new cryptanalysis results, both outputs of Claude Mythos, their (still) unreleased advanced model. The first of these results attacks a signature scheme called HAW…

0

Neil Madden @neilmadden@infosec.exchange · Jul 31, 2026

Replying to @pq1r@tech.lgbt

@pq1r@tech.lgbt sure, but for 99.9999% of cases (eg release source downloads on GitHub), the distinctions make no difference. The compression difference is marginal, and pretty much everyone just downloads and unzips the whole thing, doesn’t care about preserving file access times or whatever else. We duplicate storage, processing (eg scanning for vulns) etc to basically just satisfy the fetishes of a few purists IMO.

2

1

0

Neil Madden @neilmadden@infosec.exchange · Jul 31, 2026

Q: in 2026, do we really still need to be producing both .tar.gz AND .zip files? Surely we can just pick one?

0

1

0

Neil Madden @neilmadden@infosec.exchange · Jul 21, 2026

Fondly remembering the good old days when it was blockchain fanatics spamming the mailing lists I care about. At least they wrote their own emails.

0

Neil Madden @neilmadden@infosec.exchange · Jul 09, 2026

Replying to @Setok@attractive.space

@Setok@attractive.space 🤪

0

Neil Madden @neilmadden@infosec.exchange · Jul 09, 2026

Replying to @Xavier@infosec.exchange

@Xavier@infosec.exchange 😬

0

Neil Madden @neilmadden@infosec.exchange · Jul 09, 2026

9.28GB. 8 new emoji as the headline feature.

0

2

0

Neil Madden @neilmadden@infosec.exchange · Jul 06, 2026

Huh. I just opened search on my iphone and before I started typing anything it suggested I call my grandmother. Nice idea, but she’s been dead for 20 years.

0

Neil Madden @neilmadden@infosec.exchange · Jul 05, 2026

Replying to @NilsRenaud@m.g3l.org

@NilsRenaud@m.g3l.org There’s a whole argument going on about whether to allow a TLS key exchange which just uses ML-KEM, rather than a hybrid of ML-KEM plus elliptic curve crypto. The non-hybrid one is being pushed by NSA…

0

Neil Madden @neilmadden@infosec.exchange · Jul 04, 2026

What I don’t understand about the whole TLS ML-KEM debacle: why do it through the IETF at all, and kick off this nonsense? The registration requirements for TLS parameters are “specification required” AIUI, so any SDO can issue a spec and register the codepoints with IANA, with just a nod from the designated experts. Why pick this predictable public fight?

0

1

0

Neil Madden @neilmadden@infosec.exchange · Jul 02, 2026

Replying to @david_chisnall@infosec.exchange

@david_chisnall@infosec.exchange Tcl/Tk would like a word. (The big problem with VB from my memory was the inability to abstract, or even loop, over UI elements. If you have N items in a list and so need N copies of a widget, that was a puzzle in VB but obvious in Tk).

0

Neil Madden @neilmadden@infosec.exchange · Apr 28, 2026

Sometimes all you have is a private key and you need the corresponding public key. Today’s “fun” puzzle: implementing this for a range of key types in Java, using only standard library APIs (no BouncyCastle). Some clever/evil hacks needed.

https://codeberg.org/neilmadden/snippets/src/branch/main/PrivateToPublicKeyConverter.java

#java #cryptography

1

0

1

0

Neil Madden @neilmadden@infosec.exchange · Apr 10, 2026

Replying to @hacksilon@infosec.exchange

@hacksilon@infosec.exchange I’m also super interested in how well it generalises to non-memory-safety vulns. How load-bearing is ASan as a quality gate here, and what other classes of vulns have similar oracles?

1

0

Neil Madden @neilmadden@infosec.exchange · Apr 09, 2026

Replying to @hacksilon@infosec.exchange

@hacksilon@infosec.exchange yeah, for the OpenBSD bug they mention a “few dozen” other findings. But if they were good findings I think they would have said something about them. The fact they just say it as an aside with no elaboration suggests to me these other findings are probably a bit “meh”, but we’ll wait and see. Hopefully we’ll see the full list eventually, once disclosure has run its course.

0

2

0

Neil Madden @neilmadden@infosec.exchange · Apr 08, 2026

Replying to @neilmadden@infosec.exchange

To be honest though, with quoted figures of $10-20,000 to find each of these vulns, I don’t think they’re going after the defender market...

2

4

0

Neil Madden @neilmadden@infosec.exchange · Apr 08, 2026

I’m willing to believe that Anthropic built a better SAST. But that’s a total market of about $5B tops according to Google (some estimates seem to be just $0.5B) – it’s going to take a while to pay off their $30B Series G if they keep targeting these relatively tiny markets.

The same as with targeting developer productivity (another famously quite small market), they are focused on these markets because there are existing automated “bullshit-corrector” tools. In the case of software development, type checkers, linters, testing frameworks etc. In the case of memory corruption bugs, apparently they leant heavily on ASan to weed out the false positives.

Anyone who’s ever used a SAST on a mature code base knows that reducing false positives is the number 1 priority.

Also, in a parallel to recent articles about coding agents, finding vulnerabilities is not the bottleneck.

25

5

15

0

Neil Madden @neilmadden@infosec.exchange · Apr 01, 2026

Replying to @fooflington@infosec.exchange

@fooflington@infosec.exchange “there’s a problem with the site” for a 4xx error is really galling!

6

2

1

0

Neil Madden @neilmadden@infosec.exchange · Mar 20, 2026

Replying to @Lana@beige.party

@Lana@beige.party thinking having xy chromosomes makes you a man explains a lot about the manosphere. Neither necessary nor sufficient.

2

0

1

0

Neil Madden @neilmadden@infosec.exchange · Mar 14, 2026

It just occurred to me that if hash functions are message “digests”, then the output of such digestion is…. 💩

Brb, implementing constant-time turd comparison.

3

0

2

0

Neil Madden @neilmadden@infosec.exchange · Mar 11, 2026

Replying to @grifferz@social.bitfolk.com

@grifferz@social.bitfolk.com fascinating. And would you have to somehow switch policy depending on where you are at the time?

1

0

Neil Madden @neilmadden@infosec.exchange · Mar 01, 2026

“What I mean is that if you really want to understand something, the best way is to try and explain it to someone else. That forces you to sort it out in your own mind. And the more slow and dim-witted your pupil, the more you have to break things down into more and more simple ideas. And that’s really the essence of programming. By the time you’ve sorted out a complicated idea into little steps that even a stupid machine can deal with, you’ve certainly learned something about it yourself. The teacher usually learns more than the pupil. Isn’t that true?” — Douglas Adams

“It is not knowledge, but the act of learning, not possession, but the act of getting there which generates the greatest satisfaction.” — Carl Friedrich Gauss

“You think you KNOW when you learn, are more sure when you can write, even more when you can teach, but certain when you can program” — Alan Perlis (of course)

Why I don’t use #LLMs for #programming …

19

3

21

0

Neil Madden @neilmadden@infosec.exchange · Feb 20, 2026

Looking for vulnerabilities is the last thing I do

There's a common misconception among developers that my job, as a (application) Security Engineer, is to just search for security bugs in their code. They may well have seen junior security engineers doing this kind of thing. But, although this can be useful (and is part of the job), it's not what I focus on and it can be counterproductive. Let me explain.

http://neilmadden.blog/2026/02/20/looking-for-vulnerabilities-is-the-last-thing-i-do/

27

0

26

0

Neil Madden @neilmadden@infosec.exchange · Feb 25, 2025

Replying to @badkeys@infosec.exchange

@badkeys@infosec.exchange I’ve yet to come across a good reason for why JWK needs to represent private keys at all. I added numerous checks to prevent accidentally revealing private key material when I was at ForgeRock, but it’s such an error-prone design.

2

1

0

Neil Madden @neilmadden@infosec.exchange · Dec 18, 2023

Replying to @hanno@mastodon.social

@hanno @lambdafu looks like it’s the latter.

0

1

0

Neil Madden @neilmadden@infosec.exchange · Dec 25, 2022

Replying to @BenAveling@infosec.exchange

@BenAveling@infosec.exchange @alx@mastodon.mit.edu @HamAndChris@hachyderm.io @epixoip@infosec.exchange Depends if you are likely to be individually targeted. If you are then a predictable salt allows an attacker to precompute hashes prior to compromise, speeding up password recovery post-compromise. This is why OPAQUE goes to great lengths to hide the salt: https://datatracker.ietf.org/doc/html/draft-krawczyk-cfrg-opaque-03

1

0

Neil Madden @neilmadden@infosec.exchange · Dec 25, 2022

Replying to @neilmadden@infosec.exchange

@alx@mastodon.mit.edu @HamAndChris@hachyderm.io @epixoip@infosec.exchange also, there are two PBKDF2 processes in play here: one used for authentication to the server and an entirely separate one to derive the vault encryption key. https://www.lastpass.com/-/media/175854c49fcb489baeaa87e78579e28f.pdf

1

0

1

0

Neil Madden @neilmadden@infosec.exchange · Dec 25, 2022

Replying to @alx@mastodon.mit.edu

@alx@mastodon.mit.edu @HamAndChris@hachyderm.io @epixoip@infosec.exchange hmm… their docs say that the salt is not random, it’s your username: https://infosec.exchange/@neilmadden/109570603420250703 But those docs don’t mention anything about a client/server-side split.

0

2

0

Neil Madden

Posts

Some thoughts about Anthropic’s new cryptanalysis results

Remote instance

Media