The Bank for International Settlements (BIS), often called the central bank of central banks, has just made one of its most significant technological pivots yet. In the newly published results from the second phase of Project Aurum, its retail central bank digital currency (CBDC) research initiative, the BIS concluded that user privacy in a future digital currency system should rely on more than policy promises. It should be engineered directly into the technology, and the leading candidate is zero-knowledge proof cryptography, the same family of mathematics that already powers privacy coins and increasingly scales Ethereum.
For an institution that once faced accusations of designing digital currencies for absolute state control, the shift in tone is striking. The BIS is effectively rewriting its CBDC sales pitch, from absolute control to zero-knowledge proofs.
What Project Aurum Phase Two Actually Found
Project Aurum began in Hong Kong, a collaboration involving the Hong Kong Monetary Authority, the BIS Innovation Hub, and local financial institutions. Phase one demonstrated a working retail CBDC prototype with intermediated wallets and a token-ledger architecture. Phase two, whose results were published this week, tackled the question that has haunted every CBDC conversation: if a central bank issues digital money directly to citizens, who can see the transactions?
The conventional answer, transaction databases visible to the central bank or commercial intermediaries, is precisely what has fueled public distrust of CBDCs from Europe to the United States. Phase two of Aurum explored a different architecture, one where zero-knowledge proofs allow a payment to be verified as valid without revealing the underlying transaction details to any watching party. The payer proves they have sufficient funds and that the money has not been double-spent, without exposing their balance, identity, or spending history.
The BIS framing matters because it acknowledges the obvious: privacy assurances written in legislation can be rewritten by future legislatures. Cryptographic guarantees, by contrast, cannot be quietly weakened. If a CBDC is ever to earn public trust, the privacy model needs to survive changes in government, and only mathematics offers that durability.
Why This Matters for the Crypto Industry
Zero-knowledge proofs are not new to crypto users. Zcash pioneered large-scale deployed zk-SNARKs for private payments. zkSync, Starknet, Polygon and other Ethereum scaling projects use validity proofs to bundle thousands of transactions into compact cryptographic attestations. What is new is seeing the world’s most conservative financial institution endorse the same primitive as the foundation for sovereign digital money.
The validation runs in both directions. For years, policymakers dismissed privacy-preserving cryptography in crypto as a tool for evasion. Now the BIS is effectively conceding that privacy is a feature, not a bug, and that zero-knowledge architecture is the only realistic way to deliver it in a state-issued currency. That is a legitimization moment for the entire ZK sector, from proof-system researchers to the wallets and protocols that have shipped this technology for years.
It also raises a competitive question. If a future CBDC offers instant settlement, offline capability, and zero-knowledge privacy, what exactly distinguishes it from a well-designed privacy chain? The honest answer is issuer trust and legal tender status, not technology. The technology gap has closed, and both sides now draw from the same cryptographic toolbox.
The Trust Problem Is Not Solved Yet
Skepticism remains warranted. A zero-knowledge layer proves transactions are valid without revealing them, but someone still sets the parameters, chooses which information is revealed to regulators under warrant, and controls issuance. Cryptography constrains surveillance, it does not abolish it. Critics of CBDCs will note that the same BIS that now champions privacy has previously published work on programmability and controllability features that alarmed civil liberties groups.
The practical takeaway for readers is threefold. First, the privacy debate around digital currencies has moved from should we have privacy to how do we engineer it, and that is a genuine win for the crypto ethos. Second, ZK talent and infrastructure are about to see demand from an entirely new class of customers, central banks themselves. Third, the distinction between public crypto networks and state digital money will increasingly be decided at the level of governance and credible neutrality, because the underlying tech is converging.
Market Context
The news lands amid a soft tape for major assets. Bitcoin trades near 81,576 USD, down about 1.7 percent on the day, while Ethereum sits near 2,424 USD, down roughly 3.3 percent, and Solana around 104 USD, off about 3.4 percent as of 18:50 UTC on October 14, 2026. Privacy-oriented assets have shown relative strength this year even as the broader market churns, and institutional validation of zero-knowledge technology by the BIS adds a fundamental narrative pillar that has nothing to do with short-term price action.
Project Aurum phase two will not launch a CBDC by itself. But when the history of digital money is written, the moment the BIS stopped selling control and started selling cryptography may be remembered as the point where CBDCs and crypto stopped being opposites and started being competitors on the same technological playing field.
Disclaimer: This article is for informational purposes only and does not constitute financial advice.
funny how the same institutions that called privacy coins risky are now researching zero knowledge proofs for their own CBDC. Aurum phase two is basically an admission privacy mattered all along
intermediated wallets plus a token ledger means your bank still sees everything at the rails level. ZK here protects you from the state, not from your bank
exactly, the intermediary stays the nullifier holder in the aurum model. that is privacy from other users, your bank still maps every coffee you buy and files it happily
privacy from the state but not from your bank is the line that should be in every headline about this. the ZK hides you from the token ledger, the intermediary ledger still exists and regulators will read it
the intermediary ledger is the fine print nobody reads. your bank mapping every coffee you buy is a surveillance circle with extra steps
the intermediary ledger point is underrated. zk gives you privacy from the token ledger but your bank still files the pattern, privacy in one window and a permanent record in the other
exactly, the zk layer covers the token ledger and stops there. your issuer bank still knows you bought coffee at 8am every day this week
the coffee metadata point is the whole issue. zk on the ledger, full diary at the issuer, privacy theater at best
right, zk covers the token ledger while the bank keeps the pattern. the fix would be issuer blindness and nobody at the BIS is proposing that
issuer blindness is exactly it. blinding the token ledger while the intermediary still sees amounts and counterparties is privacy theater unless the issuing bank itself cannot link mint to redeem. Aurum phase two acknowledges the problem and then stops one step short of solving it
mint to redeem linkage is the exact gap. if the issuing bank can correlate issuance and redemption the zk layer is privacy wallpaper
mint to redeem correlation is the whole ballgame. unless the issuing bank is blinded at redemption the zk layer just decorates the surveillance
privacy from the state but not your bank is also the compromise that got the intermediary banks to sign the pilot at all
Aurum had actual retail wallet tests back in 2022, more runway than the usual BIS pdf. Still not believing any privacy claim until the audit code is public
hong kong wants the pilot crown and the hkma already has rails from phase one. if anyone ships a zk retail cbdc first its that jurisdiction
BIS doing this with the HKMA makes sense, hong kong wants the digital currency pilot crown. wonder if the aurum findings ever make it into a live retail rollout or stay a sandbox pdf forever
lived through the hkma pilot era, most of it stayed pdf forever. but aurum had actual retail tests back in 2022, this one might genuinely escape the sandbox
the irony is the tech they need already shipped in crypto. if the BIS ends up running CBDC privacy on the same zk tooling Ethereum uses, the blockchain versus cbdc framing collapses pretty fast. the real fight is about who runs the ledger, not the math
the framing collapse point is underrated. if the bis ships zk privacy on the same circuits ethereum uses, the cbdc versus crypto war becomes a ledger custody question and nothing else
BIS researching zero knowledge for its own currency while the EU rolls out traceability wallet rules tells you everything. privacy tech for the issuer, surveillance for everyone else
privacy tech for me but not for thee. they studied the math alright, studied exactly how much of it they can keep from you
studied exactly how much of it they can keep from you is the best one line summary of phase two i have read
phase one settled actual retail payments in hong kong back in 2022, thats the only reason im giving phase two the time of day. ship the zk code into a live pilot and ill believe the privacy pivot
phase one wallets running in a live pilot does buy some credibility. still waiting on anyone actually auditing the zk circuits before believing the privacy pitch
agree on the audits. zk circuit auditing is a niche skillset, maybe a handful of firms can genuinely do it, so the math being sound according to the same institution building it does not count for much yet
phase two recommending zk proofs reads like the BIS admitting cash privacy has no digital successor. you engineer what policy cannot promise