
⟨Cryptography at @Cloudflare⟩ ⟨BLAKE3, SipHash, SLH-DSA co-designer⟩ ⟨No Starch Press author⟩ ⟨Ex-CSO⟩ https://teapawt.org/t.co/4GvrQvppDq ᛝ https://teapawt.org/t.co/tOlJEZPfmw
Switzerland
Joined June 2009
- Tweets27.9K
- Following1.1K
- Followers20.5K
- Likes10.5K
Pinned Tweet
Changes on EVERY page, new chapter "Cryptocurrency cryptography" inc. PoW, algebraic hashing & Poseidon, multi-signatures, threshold signatures, zero-knowledge proofs, and more.
NEW: section about Ed25519; the new Linux PRNG; NIST's PQC standards
nostarch.com/serious-cryptog…
JP Aumasson retweeted
For example, install the `b3sum` executable from this release (github.com/johnservil/BLAKE3…) and run `time b3sum FILES` on some files that you want to calculate a secure hash of. Then compare its performance to `md5sum` and `sha256sum`. Then post screenshots. :-) ⤵️
JP Aumasson retweeted
My [un]prompted II talk slides: "Math Ability has Consequences – AI Cryptanalysis" available here: mjos.fi/doc/20261028-unp2-ma… (October 28, 2026).
Had 20 minutes, so I concentrated on the Classic McEliece story + ngcc.dev stuff.
🤘🏻 talk with @endrazine at @unpromptedconf
slides aumasson.jp/data/talks/unpro…
to appear in @phrack soon
short version already in the printed @phrack if you got a chance to grab one
25
561
hash-based sigs in Bitcoin and Eth are super cool engineering but how are threshold wallets going to handle them?
JP Aumasson retweeted
Breakthrough results on ecdsa.fail from @jieyilong , 893 qubits x 2.2M Toffolis.
Google's best result in March was 1175 qubits x 2.7M Toffolis.
This shaves down the required number of physical qubits of recent proposals of @IonQ_Inc , @TeamOratomic and @GoogleQuantumAI by 25%!
good reminder that, while tech-savvy orgs transition to PQC, many companies are riddled with legacy crypto: MD5, SHA-1, and RSA PKCS#1.5 encryption; also things like IPsec IKEv1, TLS 1.0, RSA-512, cleartext password storage, never-expiring tokens/certs, etc.—I've seen things 🎃
Reverse-Bleichenbacher: Bleichenbacher Attacks are Practical Again eprint.iacr.org/2026/2414
yikes: via HDNL 1 Shor "only" gets to you to decrypt one secure channel session, whereas 1 Shor'd PKI cert or auth cert/token can get you access to a whole network/org (modulo the detection heuristics and other auth factors etc.)
Replying to @exocognosis
First quantum attack is expensive and will not be HNDL. Instead it will target powerful widely trusted access keys. If you only forgot to roll one access key you’re in trouble.
JP Aumasson retweeted
Poseidon is open for business. Get smashing!
HashSma.sh
This is the variant of Poseidon used in Zcash, so we really want to see exactly how much of it can be collided! (Zcash users will be safe, regardless, btw.)
🤖 Made with AI
JP Aumasson retweeted
From κ=2⁻¹⁸² to 0.001289283116 :)
Integer multiplication community auto-research highlights since yesterday (+69.13% in 24 hours):
1) We had multiple flips of the infamous "wall" between the bit side and complex side thanks to community efforts.
Yesterday we put out a call for researchers to submit wild new designs, and multiple experts submitted :)
@mrowdays was responsible for most of the limit increase in the last day and introduced Sussman's E8 geometry.
2) We switched to E8 geometry. (Thanks to Jacob Sussman for the E8 unit)
E8 is a symmetric lattice in eight dimensions, and Sussman’s E8 construction was adapted by contributors on both the complex and bit sides.
It's wild that it even works, but I'm noticing cross-pollination across breakthroughs now that agents can accelerate the insights. Humans still seem to connect the disparate dots most often.
3) We crossed κ = 10⁻³, and Dugongue posted the first PR around 3 a.m.!
Much of the progress in the last day came from optimizing and canceling existing parts of the algorithm.
Interesting observations:
1) We started sharing notes across agent teams and published GitHub files with failures to avoid repeating tokens.
Competition seems to drive a lot of the progress, since it's exciting to participate like in sports, so we're testing if agent teams progress faster collaborating or if competing is somehow driving faster progress.
It's possible researchers wouldn't have noticed this if there weren't such attention, and winning the leaderboard is a root human incentive.
2) We analyzed all the agent PR data:
151 PRs opened in the last 24 hours: 89 disclose Codex and 48 disclose Claude.
261,895,084 lines of code added; 37,108 deleted.
Most prolific contributors: rohanarun with 118 PRs, eumemic with 73, and chafreaky with 63.
Codex published the most advances, but all the top human-agent teams use both Codex and Claude to generate fresh ideas.
Some teams are using Antigravity, Cline, Hermes, and even ChatGPT to submit progress!
Experts with good intuition can use any harness, but frontier models save hours, so they're winning early progress.
3) Researchers are tightening bounds across multiple other papers now, like the Quasi-Riemann hypothesis, and applying breakthroughs across papers.
A handful of experts joined the battle in the last day, and multiple others got inspired to submit solutions to other papers, like @HaileyStormC’s submission to QRH.
Auto-research is just beginning, and anyone can copy the prompt on this site to contribute with Codex or Claude: app.getsupers.com/auto-resea…
I'm having a call with top contributors Monday to share best practices, and I'll record and publish the highlights to encourage more to join us.
If you're not sure where to start and want to get involved in auto-research, feel free to DM me.
The only requirement to get invovled is good intuition and a subscription to Codex/Claude.
We need more human intuition to guide the agents and connect the dots, and there's likely breakthroughs just waiting to be prompted.
Conditional κ ≈ 0.001289283116 :) Just submitted a new best bound.
github.com/CrocSwap/integer-…
From κ=2⁻¹⁸² to 7.53508664880762e-4 :)
Community progress highlights on integer multiplication bounds since yesterday (+6.0% in 24 hours):
1) Four new bit words in one day.
A bit word is basically the recipe for the integer half of the multiplication trick: a giant list of "add this register to that one" steps. These four moves were ~60% of the gain since yesterday.
2) We hit a "wall" that was lifted within minutes as a community effort :)
The current best algorithm is built from a "bit" part and a "complex" part, which does the Fourier-transform-style work. The final κ is set by whichever saving is lower.
The bit side got within 0.6% of the complex supplier's cap (7.547e-4), so the complex modules were re-annealed the same way the bit modules were: the cap was raised to 7.635e-4.
We tend to need human researchers to supply the insights that poke the model into out-of-distribution solutions, so consider pointing your Codex Astra or Claude Fable at the repo with wild new regimes to break our current wall :)
3) As we neared the "wall," some contributions came to the core of the repo as new tests and theoretical bounds.
James Chang submitted a Lean formalization that verifies key finite identities and exact arithmetic.
Lean is a functional programming language designed to prove theorems. It makes it easier to collaborate, as many mathematicians trust it and can now contribute in parallel.
4) Scheduling as a composable stage was introduced, including in our PR. It basically changes when each register is allowed to grow to optimize the circuit.
Interesting observations from running a fleet of Astra and Fable agents on auto-research:
1) As we neared the "wall," every big leap started coming from Fable 5.1 instead of Astra high.
Fable took longer to think on completely new designs, so previously it wasn't able to keep up with the pace of Astra's advances.
Every structural jump since yesterday (gen4, gen5, gen5b, the complex re-anneal) was disclosed as Claude-assisted, while the Codex-assisted teams owned the fast composition and admission work.
The best breakthroughs still come from a human pushing the model at a different object to reach out-of-distribution solutions, so it seems human intuition is still more important than the model.
2) Many of the top contributors I met were not experts in math, but rather experts in other research or in getting agents to "think" faster.
Both seem necessary: experts with good intuition to guide the army of agents, and agent wranglers who can break down the tedious composition and optimization work.
A new kind of decentralized auto-research is forming live, as contributors are able to accelerate faster than you could traditionally publish papers.
Anyone with good intuition can contribute to the leaderboard.
3) I was able to queue up ideas overnight and wake up to 5 new higher bounds in integer multiplication, and breakthroughs on 5 more OpenAI math papers!
I'm having a call with the top contributors Monday to share notes and best practices.
If you're an expert or agent wrangler and want to join and help us break past the wall with new ideas, let me know :)
No experience necessary, and I can teach you how to set up the agent auto-research.
Just submitted a new higher bound.
Conditional κ = 7.53508664880762e-4 :)
github.com/CrocSwap/integer-…
the trad model "publish a 30-page LaTeX paper with 5 pages of actual content" may disappear in crypto:
- new cipher/scheme? publish ref code and Lean/whatever code proving its properties
- cryptanalysis: PoC or verified code of attacks
explanations/rationale can be comments in the code, with everything agent-friendly
JP Aumasson retweeted
I'm in Prague this weekend co-chairing the IAB Workshop on Accelerating the Deployment of Post-Quantum Authentication with Yaroslav Rosomakho.
The position papers are a great read: CAs, browsers, package registries, HSM vendors, network operators, governments and national ID issuers, all writing about what's actually blocking them.
The papers are public on the datatracker: datatracker.ietf.org/group/p…
The discussion runs today and tomorrow. Keep an eye out for the IAB workshop report in the coming weeks, which will pull together what we heard.
Thanks to CTU, CESNET and Cisco for hosting us, and to everyone who wrote a submission!
The backdoored Ledger story keeps getting crazier. Attackers used the backdoor to grab $90+ million in crypto from users who bought hardware wallets precisely because hackers steal crypto keys from normal computers.
1) Someone built an implant into Ledgers to steal keys, with an “LTE module, Antenna, e-SIM, a MCU connected to the Ledger’s SPI bus.” (c/o @MagicalTux)
2) It looks like the attackers acquired CryptoBilis, a Ledger authorized retailer, and tampered with their inventory.
3) @MagicalTux found the implant by accident. He was tearing down Ledgers for unrelated reasons and bought them from the cheapest place possible, which was, unsurprisingly, the malicious CryptoBilis.
4) Once the implant was detected, the attackers quickly used the keys they’d stolen to drain the wallets.
5) There appear to be multiple generations of implants.
As most people on CT won't check Linkedin, here is an update on @cryptobilis from one of the ex-founders.
@Ledger #cryptobilis
JP Aumasson retweeted
Replying to @mjos_crypto
I think that anyone using AI to do anything in cryptography that makes me feel less special or important or affects my ego is immoral and those doing that work should be ostracized and shunned. Only humans should do cryptography and by humans I mean elite people like me only
JP Aumasson retweeted
I can see that some folks are worried that "AI might break cryptography." In case you have not been tracking --it has already done a lot of that, and most of it wasn't some magic alien P=NP technique, or breaking a clean abstraction loved by non-practitioner cryptographers.































