mjos\dwez

They have been doing this for 3 years now (Finnair used to fly the route back in cold war days too.) The A350 had to be certified for 300 minute ETOPS because there is really nowhere to land up there. www.finnair.com/jp-en/bluewi...

In Japan again, and Finnair again used the "Northern Route" (I got a certificate this time!)
Recap: Russian airspace being completely blocked, this flight from Helsinki goes almost right over the north pole to reach the Bering Strait (and Alaska ATC), then turns SW to reach Tokyo.

Delving into LLM-assisted writing in biomedical publications through excess vocabulary Excess words track LLM usage in biomedical publications.

These researchers did some Delving [sic] into LLM generated text in biomedical research papers.
[ Dmitry Kobak et al. "Delving into LLM-assisted writing in biomedical publications through excess vocabulary." Sci. Adv. 11, eadt3813 (2025). science.org/doi/10.1126/... ]

I mean, those things basically require no action at all. An intern can compose such a list of product categories, and TLS 1.3 by itself has nothing to do with PQC.

Some seem to have interpreted the Trump June 6 "cybersecurity diff" EO to set some kind of deadline on PQC but no. This EO imposes a deadline for a PQC product categories (?) and, uh, TLS 1.3.
(However those PQC deadlines still exists in other places.) www.whitehouse.gov/presidential...

Abstract. A multi-message multi-recipient Public Key Encryption (mmPKE) enables batch encryption of multiple messages for multiple independent recipients in one go, significantly reducing costs, particularly bandwidth, compared to the trivial solution of encrypting each message individually. This capability is especially critical in the post-quantum setting, where ciphertext length is typically significantly larger than the corresponding plaintext. In this work, we first observe that the generic construction of mmPKE from reproducible PKE proposed by Bellare et al. (PKC ’03) does not apply in the lattice-based setting because existing lattice-based PKE schemes do not fit the notion of reproducible PKE. To this end, we first extend their construction by proposing a new variant of PKE, named extended reproducible PKE (XR-PKE), which enables the reproduction of ciphertexts via additional hints. However, standard lattice-based PKE schemes, such as Kyber (EuroS&P ’18), do not readily satisfy the XR PKE requirements. To construct XR-PKE from lattices, we introduce a novel technique for precisely estimating the impact of such hints on the ciphertext security while also establishing suitable parameters. This enables us to instantiate the first CPA-secure mmPKE and Multi-Key Encapsulation Mechanism (mmKEM) from the standard Module Learning with Errors (MLWE) lattice assumption, named mmCipher-PKE and mmCipher-KEM, respectively. We then extend our works to the identity-based setting and construct the first mmIBE and mmIB-KEM schemes. As a bonus contribution, we explore generic constructions of adaptively secure mmPKE, achieving security against adaptive corruption and chosen-ciphertext attacks. We also provide an efficient implementation and thorough evaluation of the practical performance of our mmCipher. Our results show that mmCipher provides significant bandwidth and computational savings in practice, compared to the state-of-the-art. For example, for 1024 recipients, our mmCipher-KEM achieves a 23~45 times reduction in bandwidth overhead, reaching within 4~9% of the plaintext length (near optimal bandwidth), while also offering a 3~5 times reduction in computational cost.

Image showing part 2 of abstract.

Image showing part 3 of abstract.

Post-Quantum Multi-Message Public Key Encryption from Extended Reproducible PKE (Hongxiao Wang, Ron Steinfeld, Markku-Juhani O. Saarinen, Muhammed F. Esgin, Siu-Ming Yiu) ia.cr/2025/1000

btw hardwear.io was great! recommended

the slides are at mjos.fi/doc/20250530...

Post-Quantum Multi-Message Public Key Encryption from Extended Reproducible PKE A multi-message multi-recipient Public Key Encryption (mmPKE) enables batch encryption of multiple messages for multiple independent recipients in one go, significantly reducing costs, particularly b...

new paper eprint.iacr.org/2025/1000
artifact is at github.com/ml-kem/mmcip...

"bitcoin style encryption"

The talk is on Friday at 11:15 (hardwear.io/usa-2025/sch...) I released some of the tooling used to generate rough pre-silicon "toggle" traces github.com/ml-dsa/abr-sim

NIPRnet and SIPRnet PKIs were RSA-2048 still in December 2022 according to this earlier DoD directive. Note that there is no RSA or elliptic curves of any length in the "new" CNSSP-15, so this seems completely void.

Btw CNSSP-15 is publicly available: "Use of Public Standards for Secure Information Sharing" (Version: March 2025)
Nothing really new here, but "CNSSP-15" now seems like the most appropriate thing to cite wrt CNSA 2.0. www.cnss.gov/CNSS/issuanc...

To give you the idea: ".. yeah I did this really pretty complicated electromagnetic side-channel lab stuff but also noticed that whole thing just cracks open if you stick garbage into this one port fast enough.." It's the garbage-attack thing that hackers will exploit.

Why Why

Having fun making a talk for a hands-on security conf.. In academic research we often ignore simple implementation bugs, API design flaws, and other "trivial" ways to break crypto systems. But for a real-world security talk such findings are *gold*.. 😅 hardwear.io/usa-2025/spe...

I've been wanting to do the Belfast City Marathon ever since I lived here 2015-16. Had the opportunity to do the run this weekend. No official timing yet, but it will be something like 4:06. Yes, they let runners choose their bib numbers for £5, and I couldn't resist 😉