Verona, the verified intelligence layer for AI and enterprises, launched Fermi on October 28, 2026, the v32 upgrade to its mainnet that brings live post-quantum signatures to every account on the network. Users can attach a post-quantum authenticator to an existing account and revoke the old classical key, all without changing their address, moving funds, or migrating to a new account, a design Verona says solves the hardest part of the quantum transition: not the cryptography itself, but the migration.
What Fermi Ships
Fermi adds two NIST-standard signature schemes as native account authenticators. The first is ML-DSA-65, the FIPS 204 lattice-based standard, available as a general account authenticator. The second is SLH-DSA-128s, the FIPS 205 hash-based standard, aimed at treasuries and cold vault accounts where conservative, long-lived security matters most. A new PQ-required account policy can disable classical keys entirely, or force them to co-sign alongside a post-quantum key, and migration is offered as a one-click operation in the Verona software kit and dashboard.
The timing, Verona argues, is about lead time rather than panic. NIST plans to deprecate the elliptic curve signatures that secure nearly every blockchain after 2030 and disallow them after 2035, while Google has moved its own post-quantum migration deadline to 2029. In March, Google Quantum AI published estimates suggesting a machine with fewer than half a million physical qubits could break secp256k1, the curve behind Bitcoin and Ethereum, in minutes, a nearly twentyfold drop from prior estimates. No such machine exists today, Verona notes, which is precisely why it chose to migrate while the lead time still exists.
Why The Account Design Matters
Most blockchains derive an address from a single public key, so when that key’s algorithm breaks, users must move funds to a new address. Verona says that design leaves an estimated one third of Bitcoin supply sitting in accounts with exposed keys, and forces Ethereum’s quantum roadmap through new account standards. Verona built its accounts differently from the start: every account runs as a governance-deployed smart contract holding a set of pluggable authenticators, today spanning secp256k1, passkeys, Ethereum wallets, JWT and ZKEmail.
Adding ML-DSA uses the same operation as adding a passkey, because the contract owns the address rather than the key. Cryptography can therefore change underneath an account while identity, history and balances stay put, which is what makes Fermi a key swap rather than a migration. The company is explicit that Fermi will not end the cat-and-mouse game between quantum computing and cryptography: it runs lattice and hash-based keys side by side, so if either assumption fails, users can move to the alternative.
The Bet, And What Comes Next
“Quantum resistance is a migration problem, not a technical problem,” said Anthony Anzalone, founder and chief executive of Verona. “Every other chain has to figure out how to move billions of dollars to new addresses. On Verona you add a key and remove a key. We built accounts this way years ago because we knew cryptography would keep changing, and Fermi is the first time that design pays off at network scale.”
Fermi’s second phase will migrate validator keys to ML-DSA-65 through the Cosmos SDK v0.55 and CometBFT v0.40 line, moving consensus itself to post-quantum signatures once two-thirds of validator voting power rotates. Verona will add FN-DSA, known as Falcon, as an authenticator once NIST finalises FIPS 206. Convenience authenticators that depend on third-party infrastructure, including JWT and ZKEmail, stay classical until identity providers and proof systems finish their own migrations, and they cannot satisfy a PQ-required policy.
Why It Matters
This matters because the quantum threat is usually discussed as a distant cryptography problem, while the real cost is operational: moving billions of dollars of value to new addresses, with all the lost keys, broken integrations and stranded accounts that implies. Verona’s answer reframes the problem as account architecture, and if its approach works at network scale, it becomes a template other chains may have to copy, much as quantum computing’s push into materials discovery is forcing adjacent industries to plan around quantum timelines rather than react to them.
The sceptical view is that Fermi is still only phase one: user accounts are protected, but consensus itself waits for validator rotation, and the convenience authenticators most people actually use remain classical. Post-quantum security is only as strong as the weakest key that can still sign, a discipline that also applies to machine-checked proofs for code and chips and to safety benchmarks that expose where AI systems fail. Fermi’s value will be proven by how many accounts actually rotate, not by how many can.
Background
Verona, built by Burnt Labs and formerly known as XION, describes itself as the verified intelligence layer for AI, enterprises and consumers, turning verified real-world facts into a programmable primitive, and Fermi is the v32 upgrade to its mainnet.
Source: Verona announcement
0 Comments