Yesterday, news broke that Dominion Voting Systems was sold to a new company, Liberty Vote. Dominion, the second-largest voting systems vendor in the US that currently tabulates 1 in 5 American votes, is no more. www.voting.works/news/public-...

Discord user IDs getting leaked is the entirely predictable consequence of requiring platforms to do age verification. That data never goes away, it spreads. In this case, into appeals in a breached customer support database. And predictably, it can get worse. www.404media.co/the-discord-...

The next COED Industry Day is going to be held in Milan.... coedcon.github.io/index.html Going to be a super cool mix of speakers, from Bosch, Nokia Bell Labs, Cybernetica, Roseman Labs and Zama

My colleagues from Cybernetica and I decided to make another variant of Smart-ID / SplitKey protocol. Coppercloud is a two-party blind RSA signing with private key being split between a user and a server. Perfect for the future of privacy-preserving digital wallets! 🌐✨

What is chat control? Good video explainer developed by @carmelatroncoso.bsky.social and team at Max Planck Institute for Security and Privacy #chatcontrol www.youtube.com/watch?v=-y2O...

Tehisintellekt aitab neid, kes seda kontrollida suudavad www.err.ee/1609806123/d...

A lot of generally principled people have been riding the “I do this because I believe in the mission of stopping bad guys” and I think those principles are going to be tested soon.

@lilimilani.bsky.social Would there be value in validating this study on Estonian biobank data? www.nature.com/articles/s41...

The EU Parliament has published a new proposal for Chat Control to mass-surveil all digital communication in Europe. The proposal is ineffective, weakens secure communication, and violates basic human privacy. This must be stopped immediately. #ChatControl csa-scientist-open-letter.org/Sep2025

I'm sympathetic, but I do think it is charming how teh lawyerz believe that you can legislate your way out of a threat model. "No one would ever do that, it's illegal."

Check out the new open letter signed by more than 450 scientists who have serious concerns on the latest (July 2025) version of the chat control proposal. csa-scientist-open-letter.org/Sep2025

Bart Preneel warns EU’s #chatcontrol is dangerous & ineffective: criminals will evade scanners, millions of false positives expected, and end-to-end encryption for all could collapse, threatening privacy, journalists & activists. #Privacy #Encryption #EU www.demorgen.be/nieuws/meer-...

Teaduste akadeemia küberturvalisuse komisjon avaldas 2024-2025 hooajal tehtud riskianalüüsid. koodivaramu.eesti.ee/riigi-valimi...

Really speaking to US academics here: rumor I heard is that DARPA appears to be “just not starting new (unclassified) projects.”

Teaduste akadeemia küberturvalisuse komisjoni augusti istungi protokoll on avaldatud. www.akadeemia.ee/akadeemia/no...

Usaldusest ja usaldatavusest 2025 päevakava ja esinejate nimekiri on nüüd lõplik. www.akadeemia.ee/konverents-u... Palun registreerige! Teemasid: Platvormivõim: kelle huvid ja väärtused meid juhivad? Käideldavus või kontroll? Millises ulatuses pannakse Eesti riik pilve?

New science from Cybernetica: Proving vote correctness in the IVXV internet voting system by Taaniel Kraavi and Jan Willemson www.nature.com/articles/s41...

Varajane eelteade - mõned esinejad on veel puudu - aga Teaduste akadeemia komisjon konverents Usaldusest ja usaldatavusest tuleb jälle! Registreerida juba saab. www.akadeemia.ee/konverents-u...

Excellent explanation of what chat control would mean and why we should not have it. Estonia uses so much cryptography at a national level that weakening by law it will cause legal issues for our government. chatcontrol.dk/en/

And more ZK: Novel Approximations of Elementary Functions in Zero-Knowledge Proofs link.springer.com/chapter/10.1...

And a paper on how to industrialise synthetic data generation. SynthGuard: Redefining Synthetic Data Generation with a Scalable and Privacy-Preserving Workflow Framework link.springer.com/chapter/10.1...

Zero-Knowledge Proof-of-Location Protocols for Vehicle Subsidies and Taxation Compliance link.springer.com/chapter/10.1... Some really work from back in the DARPA SIEVE program that we got to present at the DARPA Demo Day at the Pentagon a while ago.

A milestone for open AI: ETH Zurich and EPFL will release a fully open multilingual language model, trained on CSCS’s “Alps” supercomputer. Built for the public good, this model promotes transparency, inclusion, and innovation. Read more:

New paper written with my colleagues Jelizaveta Vakarjuk and Peeter Laud from Cybernetica just got accepted to ProvSec2025 🥳 In this paper, we aim to combine server-supported signatures together with fail-stop properties. P.S. yes, protocol’s name is inspired by Mistborn series

Müürilehes: massiline jälgimine (pildistaks kõiki alati ja igal pool, igaks juhuks?) ei ole mitte ainult ebaproportsionaalne ja tehniliselt tehtamatu, vaid õõnestab ka väärtusi ja usaldust, millele e-Eesti ehitatud on. www.muurileht.ee/kas-tahad-et... #privaatsus #usaldus #suurvend

Want Big Brother to read your messages? For Müürileht 🇪🇪 on why mass surveillance and law enforcement exclusive access to digital comms are disproportionate, technically unfeasible, and fundamentally undermine the values, trust digital societies depend on. www.muurileht.ee/kas-tahad-et... #privacy

The last point is spot on. The world has been systemically lied about what quantum computers will solve. I understand, why. And - the things a QC can do are pretty impressive. But do have a look at the rest of Scott's slides, linked in the original thread.

Going on vacation today.

Kas EL suudab ohjeldada infokaost ja manipuleerimist ühismeedias? - ise küsisin, ise katsusin vastata Arenguseire Keskuse palvel. 🔗 arenguseire.ee/pikksilm/kas... või rito.riigikogu.ee/nr-51/kas-el... #EU #DigitalPolicy #Disinformation #SocialMedia #Democracy #InformationDisorder

AI hariduses kasutamise paradoksist TI-hüppe kontekstis opleht.ee/2025/06/ti-k...

This is how you build secure systems: public source code, solid theoretical foundation, and third-party audits. Impressive. (Note: maybe we'll find flaws in this! And then they'll get fixed, thanks to transparency. That's also how secure systems work.) www.theguardian.com/gnm-press-of...

Decentralized Proof-of-Location systems for trust, scalability, and privacy in digital societies New paper in Nature Scientific Reports from Eduardo Brito and Liina Kamm from Cybernetica, Amnir and Ulrich from U of Tartu. www.nature.com/articles/s41...

Last night I went to see Mission Impossible: Final Reckoning, where a rogue AI takes over the entire US nuclear arsenal, and all I could think was: this shit wouldn’t have happened if they’d published ISO 19790:2025 for free.

Cybernetica is looking for a General Counsel to join our team of awesome technology-aware lawyers. Contracting is a surprisingly important part of pushing new cybersecurity, data protection and AI technologies. I have a lot of respect for these people.

Multi-Party Distributed Point Functions with Polylogarithmic Key Size from Invariants of Matrices (Toomas Krips, Pille Pullonen-Raudvere) ia.cr/2025/978

Me and Pille Pullonen-Raudvere uploaded a paper to ePrint called Multi-Party Distributed Point Functions with Polylogarithmic Key Size from Invariants of Matrices. (eprint.iacr.org/2025/978). It is a candidate solution for multiparty distributed point functions with a polylogarithmic cost. (1/12)

Cybernetica got its research evaluation report this week, signed by the ministry. Evaluation happens every 7 years and Cybernetica's work is considered by an international committee, under the same rules as university departments. We received quite the positive evaluation. etag.ee/wp-content/u...

Is MPC pseudonymisation, anonymisation, joint controllership, de-identification? The panel at the Real World MPC workshop had some of the best privacy tech experts in Europe to explain. Including Triin, who does the explaining work with customers for a living. vimeo.com/1078973800/4...

The Github MCP attack (AI agents looking at text data and then doing things based on them) means that we are now in a future Von Neumann model where every textual data field in every database is suddenly executable even - if you didn't plan it. This makes cybersecurity attack surfaces ridiculous.

The videos from Future Cryptography are like a good LP (music album). The songs in order tell a three-part tragicomedy on good EU intentions on improving cybersecurity, resources for actually doing it, tools that we have for it. Listen to it here: futurecryptography.eu/2025-present...

Convert your synthetic data generation code into a pipeline with SynthGuard. github.com/SynthGuard This has been a strong effort in the LAGO, TEADAL and EXAI projects at Cybernetica. Thanks to everyone involved!

Ahem.

It's time to start sharing results from the work that Cybernetica has done on MPC for Eurostat, in the JOCONDE project. First - we thank Eurostat for all the effort they have put into understanding MPC and putting together requirements. You find the concept on the website: cros.ec.europa.eu/joconde

"Why EU encryption policy needs technical and civil society input" www.helpnetsecurity.com/2025/05/19/b...

Wrote a thing about youth unemployment and ai. Made me sad stateofthefuture.substack.com/p/young-peop...

"Absolute Mode". For when you're tired of sycophancy.

(This one's in Estonian and on cryptography standardisation and certification, titled "Should you trust mathematicians with your secrets?") Kas matemaatikutele tasub oma saladusi usaldada? arvamus.postimees.ee/8245935/dan-...

How does WhatsApp implement encrypted group chats? And are they secure? @malb.bsky.social, @bedow.bsky.social and myself were keen to figure this out. After two years of reverse-engineering, analysis and a few too many proofs, I presented our work at Eurocrypt earlier today. So, what did we learn?

EDRi, in a group of 39 organisations and 43 experts, published an open letter today to call for a scientific evidence-based approach to encryption. We ask for meaningful participation to safeguard cybersecurity and fundamental rights in the process. edri.org/our-work/tec...