BTC/USD $68,420 +2.8%
ETH/USD $3,540 +1.4%
SOL/USD $142.80 -0.6%
BNB/USD $605.20 +0.9%
XRP/USD $0.62 -1.2%
DOGE/USD $0.18 +5.4%
BTC/USD $68,420 +2.8%
ETH/USD $3,540 +1.4%
SOL/USD $142.80 -0.6%
BNB/USD $605.20 +0.9%
XRP/USD $0.62 -1.2%
DOGE/USD $0.18 +5.4%
Guides

Stellar’s Quantum Fix Leaves Regular XLM Wallets Behind

Key Takeaways Validators vote on Protocol 30 November 23. Smart accounts could opt into ML-DSA approval. Recovery keys can still override spending rules. Classic accounts and transactions sta

AnonymousCryptoCompass newsroom
November 22, 2026
6 min read
NEWS
Stellar’s Quantum Fix Leaves Regular XLM Wallets Behind
CryptoCompass editorial visual for guides coverage.

Key Takeaways

  • Validators vote on Protocol 30 November 23.
  • Smart accounts could opt into ML-DSA approval.
  • Recovery keys can still override spending rules.
  • Classic accounts and transactions stay on Ed25519.

Stellar’s official upgrade schedule sets the mainnet validator vote for November 23 at 17:00 UTC, and the quantum-related change is one of three proposals in the package.

CAP-0087 adds three host functions to Soroban, Stellar’s smart-contract environment, one for each ML-DSA parameter set: ML-DSA-44, ML-DSA-65 and ML-DSA-87. Verification runs inside Stellar Core, so a contract calls a shared, network-maintained implementation instead of shipping its own.

NIST published ML-DSA as FIPS 204 in August 2024 and considers it secure even against an attacker who holds a large-scale quantum computer.

The attack it guards against is specific. Ed25519, the scheme behind every Stellar account today, is safe because nobody can work backwards from a public key to its private key. A sufficiently large quantum computer running Shor’s algorithm could. On Stellar, that public key is not hidden: a classic G-address is an encoding of it, so every account’s key has been public since the account was created. No machine capable of the attack is publicly known to exist, but nothing can make an exposed key private again. The only defence is to move funds under a different kind of key.

That defence has a size cost. The three ML-DSA variants step up in security level and in length, and even the smallest signature is roughly 38 times longer than an Ed25519 signature. Soroban charges contracts for the resources they consume, so the extra weight is a cost question every wallet team verifying ML-DSA on each payment has to work through.

Signature and key sizes Scheme NIST category Public key Signature Ed25519 Not quantum-resistant 32 B 64 B ML-DSA-44 2 1,312 B 2,420 B ML-DSA-65 3 1,952 B 3,309 B ML-DSA-87 5 2,592 B 4,627 B

NIST categories run from 1 to 5; a higher number means a wider safety margin. Sources: NIST FIPS 204 and IETF RFC 8032.

None of this reaches a user unless a smart account calls it. A smart account is a contract that holds funds and sets its own definition of valid approval, and its developers decide whether that definition includes an ML-DSA signature.

How a Payment Clears Two Separate Checks

Stellar’s authorization framework lets a contract account apply its own authentication rules, and it separates the account whose funds move from the account that submits the transaction. A signature proves only that a key approved certain data, as Coindoo’s guide to what a digital signature verifies sets out. Whether that approval is enough to release money is the contract’s decision.

Say a business keeps tokens in a smart account whose policy demands an ML-DSA signature for any transfer, and it needs to pay a supplier. Its wallet builds the spending instructions and signs them with the business’s ML-DSA key. A separate submitting account wraps that contract call in a network transaction and signs it with Ed25519, as every Stellar transaction is signed today. When the transaction executes, the smart account checks the ML-DSA signature against its policy before any tokens leave.

The Ed25519 layer is therefore not a back door into those funds. An attacker who forged the submitter’s Ed25519 signature could get a transaction accepted by the network, but the contract would still reject a transfer that lacks the business’s ML-DSA approval. That is what lets a business protect contract-held funds now instead of waiting for the network’s own signatures to change, provided the contract offers no weaker route to the same money.

Where a Wallet’s Protection Can Leak

Supporting ML-DSA and requiring it are separate settings. A wallet can list the new scheme among its signer options while leaving funds reachable without it, and the weak point is rarely the spending rule users see.

The first gap is an either-or signer list. If a policy accepts an ML-DSA signature or an older Ed25519 key, the older key alone still moves funds, and an attacker simply targets that one. Only a policy that demands ML-DSA on every spending path, alone or alongside other signers, keeps it in the chain of approval.

The second is recovery. Many smart-wallet designs let a guardian, a backup key or a service provider replace lost signers. If that recovery authority signs with Ed25519, whoever controls it can install a new key and spend under the new rules. The ML-DSA requirement then holds only until someone with recovery power rewrites it.

The third is the contract itself. Soroban contracts can be built to be upgradable, and an administrator key that can replace the contract code can replace the policy along with it. A wallet’s quantum resistance is only as strong as the weakest key that can change its rules.

READ MORE: Gold’s Crash Weakens the Case Against Bitcoin

What Bell Leaves on Ed25519

Everything outside smart-account authorization stays as it is under Bell. CAP-0087 states that the network’s transaction signatures, account master keys and the overlay remain Ed25519-based. The overlay is the channel validators use to exchange the messages through which they agree on ledger history.

That covers classic accounts, the standard G-address accounts Stellar wallets create by default. Their protection gets its own stage in Stellar’s Quantum Preparedness Plan, which targets native quantum-safe signers for classic accounts in 2027 and lets holders keep their existing addresses. Until that stage ships, an XLM balance in a classic account is protected by exactly the cryptography it has today.

Protocol 30 also carries unrelated compatibility work: W-prefixed muxed contract addresses and a changed ledger header format, which wallets, exchanges and indexers must support. The upgrade guide says existing contracts keep running unchanged, and installing the new software leaves every account’s signing rules where they were.

What Would Show Anyone Is Using It

If validators approve Bell on November 23, the verifier goes live. Three developments would show whether anyone relies on it.

Signals to watch after the vote

  1. A published policy. A named wallet or custodian stating which actions require ML-DSA, which keys can recover the account and who can upgrade the contract.
  2. An audit of the whole wallet. A review of the full permission structure, beyond confirming that the verifier returns correct results.
  3. Progress on the 2027 stage. Native quantum-safe signers for classic accounts are the step that would reach holders who never use a smart account.

Until the first of those appears, a wallet’s ML-DSA support is a claim to test against one question: can anything other than an ML-DSA signature move the funds, or change who is allowed to move them?

The post Stellar’s Quantum Fix Leaves Regular XLM Wallets Behind appeared first on Coindoo.