Bank Blockchain Privacy: The Layer That Pilots Keep Hitting

JPMorgan Chase, Citigroup, Bank of America and Wells Fargo are testing a shared tokenized deposit network. SWIFT has signed up more than 40 banks for its own version. Parallel pilots for tokenized money-market funds, tokenized deposits and stablecoins are running at dozens of global lenders.

The path from pilot to production keeps hitting the same obstacle: privacy. Ravi Chamria, CEO of blockchain infrastructure firm Zeeve Inc., argues the problem is not a choice between public and private chains. It is a design gap that runs from the ledger to the application layer.

"The chain may be private, but the data may still not be private across the full infrastructure," Chamria wrote in a recent analysis.

Banks have moved to permissioned and consortium networks as a first step. Those networks control who can join and which nodes validate transactions. Network-level access is not the same as data privacy. A validator on a private chain can still see all tokenized deposit movements. One bank can infer another's client flows. A client may see its business activity exposed to other participants.

Chamria calls this the gap between network-level privacy and "system-wide privacy." Sensitive information can leak through validators, RPC endpoints, observability tools and cross-chain messaging, even on a permissioned chain.

He divides the requirement into three areas. Transaction privacy ensures token amounts, balances and counterparties are not exposed across the network. Workflow privacy covers the processes around a transaction, such as collateral management, treasury movement and settlement, that can reveal intent even when the token transfer is hidden. Both need to be supported by a modular privacy layer that works with whatever chain the bank chooses. That layer handles confidential transfers, private smart contracts, zero-knowledge proofs, selective disclosure and compliance rules.

Zero-knowledge proofs let banks prove validity and eligibility without exposing the underlying commercial data. Chamria said they are insufficient alone. They must work with identity, governance, key management and auditability. "Disclosure should happen only when required, to the right party and for the right reason," he wrote.

Chamria outlined three broad setups. A private network gives a bank full control but can create an inefficient island with poor interoperability. Institutional public networks allow faster starts and offer some privacy mechanisms, but customization is limited. Consortium networks face the toughest privacy challenge because interoperability multiplies the number of parties that can infer sensitive activity. Encrypted token standards and permissioned access in Fabric or Besu-like networks help, but basic controls are not enough when multiple banks are involved.

Banks in consortium settings need separate privacy tools that mask tokens, identities, contracts, ledger data and workflow information. Chamria suggested a phased approach: start with confidential asset transfers and access controls, then protect the processes around the transaction, finally connect privacy with identity, compliance rules and disclosure governance with defined access paths for regulators and auditors.

"Without system-wide privacy, the next wave of bank blockchain adoption may repeat the same pattern as earlier pilots that never reached production," he wrote. "In banking, confidentiality is one of the conditions that determines whether blockchain-based financial infrastructure will work at all."

AlphaScala's Alpha Score rates the largest U.S. banks pushing into tokenization as Moderate, with BAC and JPM at 60 out of 100 and WFC at 56. The industry is moving ahead, but Chamria's diagnosis suggests the real technology hurdle is less about throughput or cost and more about convincing regulators and peers that sensitive flows stay locked down.