What Post-Quantum Cryptography Protects, and What It Doesn’t?

Whenever a major theft makes the news, people ask whether stronger cryptography would have stopped it. For post-quantum cryptography, the answer is no, and that surprises a lot of people. It has one job: keeping digital signatures and key exchanges safe from a future quantum computer. 

A lot of crypto security trouble sits outside that job, and users who mix the two up end up worrying about the wrong thing. Anyone trying to judge quantum security in crypto needs to know where the line fails, and how real the quantum threat is today.

Quantum Computer Cannot Steal Crypto

Can quantum computers steal crypto? No. The estimates of what such a machine would need keep falling, and that is why the topic keeps coming back. In March 2026, researchers at Google Quantum AI, with co-authors from the Ethereum Foundation and Stanford, published a paper that estimates that breaking the 256-bit elliptic curve behind Bitcoin and Ethereum would take less than 500,000 physical qubits on a superconducting machine. This is 20 times less than the earlier estimate it compares against. The circuits would run for about 18 to 23 minutes, or around 9 minutes once the first phase is precomputed (PostQuantum.com). The paper does not claim such a machine exists or say when one will, and the hardware it assumes has not been demonstrated at that scale.

In 2019, Craig Gidney and Martin Ekerå calculated that factoring a 2048-bit RSA key would take 20 million noisy qubits. Gidney’s 2025 paper brought the figure under one million. Nobody knows when a large enough machine will exist, and moving a blockchain, its wallets, and its exchanges to new cryptography takes years, so preparation has to begin before the machine does.

Most of the exposure sits in the signatures. Bitcoin and Ethereum both sign with ECDSA over the secp256k1 curve, and Shor’s algorithm can turn a public key into the private key behind it. An Ethereum address is a hash of the public key, staying hidden until the account sends its first transaction. After that, it sits on the ledger for good. 

Hash functions and symmetric ciphers hold up better. A quantum computer only halves their effective strength, so keccak-256 and AES-256 keep 128 bits of security. Network traffic has its own version of the problem. An attacker records encrypted traffic now and reads it once a quantum computer exists.

What Post-Quantum Cryptography Protects

NIST published its first three post-quantum standards in 2024 after a selection process that was initiated in 2016.

On a blockchain, post-quantum cryptography can sit in three places. Account signatures decide who may spend. Validator signatures decide which blocks count. And the exchange between nodes decides who can read their traffic. Cover all three and a quantum computer has no easy way in at any of them. If you’re curious to know how attackers go after the older signatures, have a look at this post in which we discussed if AI could break ECDSA or ML-DSA.

One Recent Incident Shows Where Signatures Stop Helping

Binance reported an estimated $6.3 billion incident on 29 October. Its early findings say it may have started with a compromised DevOps engineer’s laptop, which the attacker is reported to have used to reach a service container and submit unauthorised withdrawals. Binance also says it found no evidence of a leaked wallet, and the investigation is still open (CoinDesk, CyberInsider).

Nothing in those reports involves a forged signature or a quantum computer, and the facts may change once the post-mortem is published. Even so, the case shows the split clearly. A signature algorithm, post-quantum or not, answers one question: can someone forge this signature? Whether a withdrawal request should be signed at all is decided somewhere else, by limits, allowlists, approval flows and isolated signing environments. Replace every signature on that system with ML-DSA and the path the reports describe would still be there.

What It Does Not Cover

A Valid Signature Doesn’t Make a Transaction Safe

The 2016 attack on The DAO exploited a reentrancy flaw, a bug that let the attacker call a withdrawal function again before the balance updated. Every transaction in it was properly signed. Signature schemes verify who sent a transaction. They say nothing about whether the contract receiving it works accurately. For ordinary users, the bigger day-to-day threats are phishing, malware, social engineering, and poor wallet security. Someone who approves a malicious transaction or types a seed phrase into a fake site loses funds under ECDSA, ML-DSA, or anything else. A seed phrase photographed and left in a cloud album is just as exposed.

Even Sound Standards Get Implemented Badly

ARMchain’s own whitepaper reports that negative testing found and fixed a crafted-calldata panic in the ABI decoder used by its precompiles, a flaw that would have stopped every node from executing a block. It was caught before any public network existed. The same paper mentions that no external security audit has been performed; however, an audit is planned before mainnet. Good algorithms still need tested code.

Old Accounts Stay Exposed After an Upgrade

A chain can add post-quantum signatures and keep accepting the old ones for compatibility. ARMchain does exactly that. Its whitepaper says accounts that use standard Ethereum transaction types are “exactly as exposed as on Ethereum,” and passkey accounts, which use the P-256 curve, are classical too. On any chain that claims quantum protection, check which signature your account uses.

Quantum-Proof Is a Marketing Label

Products sold as quantum-proof deserve suspicion. Rushing into an unfamiliar wallet or a hurried migration can add risk now for a threat that is still years away. No signature scheme comes with a proof that it can’t be broken. ML-DSA, ML-KEM and SLH-DSA became standards because they survived scrutiny. Quantum-safe and quantum-resistant describe algorithms built to withstand known quantum attacks. Quantum-proof promises more than anyone can show.

NIST standardised a hash-based scheme alongside the lattice-based ones for exactly this reason. If a flaw turns up in one family of maths, the other family doesn’t share it.

What Can Be Done Now

For most people, quantum security looks a lot like good security in general.

• Keep the seed phrase offline and never type it into a site, app, or chat that asks for it.
• An external wallet is recommended for larger holdings, and it is important to read the transaction details on the device screen instead of trusting the browser.
• Avoid reusing addresses. On chains like Bitcoin, reuse exposes your public key sooner.
• Install wallet software only from the official site and update it.

Follow your chain’s upgrade announcements and migrate when its guidance says to. Expect scammers to use quantum fears, so treat any message telling you to move funds to a “quantum-safe wallet” as a red flag.

Post-Quantum Signatures at the Core of ARMchain

ARMchain works on the signature layer. Default accounts and validators sign with ML-DSA-44, the FIPS 204 algorithm, and nodes connect using ML-KEM-768, the FIPS 203 algorithm. The whitepaper keeps its claim narrow: no asset, validator, or node depends on a public-key problem a quantum computer solves, unless its owner chose a classical scheme for compatibility. Phishing, laptop compromises, and contract bugs sit outside that claim.

The devnet is open to developers who write Solidity. The ARMchain documentation covers the SDK and account types, and the whitepaper lists what is measured and what is still unfinished.

FAQs

Is it possible to steal crypto from a quantum computer?

No. Current machines are far from the scale that the published estimates call for, and those estimates themselves rest on hardware nobody has built yet. The thefts you read about in the news come from a compromised system.

Does post-quantum cryptography prevent hacks?

No, not the hacks that make the news. It only covers attacks where a quantum computer forges a signature or decrypts recorded traffic. Phishing, stolen credentials, and contract bugs are not affected by it.

Is post-quantum cryptography unbreakable?

No scheme can promise that. NIST’s standards came out of public review, and their security rests on hard maths problems. That is also why a hash-based backup exists.

Should I move my crypto because of quantum computers?

Most users need to change very little for now. Look after your keys, avoid unnecessary address reuse, and follow your chain’s upgrade guidance.