The Bank for International Settlements (BIS) has concluded that digital cash can achieve "cash-like privacy," but warned in its Project Aurum Phase 2 report, published Tuesday, that the underlying technology stack is not yet "production ready."

The report identifies bank data breaches as a key remaining risk. BIS found that compromising a single bank may expose that bank's customer data, but would not expose customer identities from other counterparties. A validator-level compromise alone would not be enough to unmask end users, according to the report.

What did Project Aurum Phase 2 actually build?

The phase implements a privacy-preserving digital cash prototype built on BIS's existing Aurum codebase. The design deliberately sits between two poles: complete anonymity and full transaction transparency. BIS describes this as a balanced approach, one intended to preserve cash-like discretion for users while keeping enough structure for the system to function.

The work is squarely aimed at central bank digital currencies (CBDCs). The report explores the technical features needed to establish a privacy-preserving CBDC, a category that has drawn sustained criticism over its potential for financial surveillance.

That criticism isn't hypothetical. In 2023, Brazil's central bank published the source code for its CBDC pilot, and within days observers spotted surveillance and control mechanisms embedded in the code. Those mechanisms would have allowed the central bank to freeze or reduce user funds held in CBDC wallets.

The political headwinds in the United States are just as concrete. In January, President Donald Trump banned federal agencies from issuing a CBDC, citing financial stability and privacy concerns. Republican lawmakers have since pushed for a permanent statutory ban as a House vote approached.

Why the breach finding matters for banks

The BIS findings carry operational weight for commercial banks. If a single bank's compromise exposes only its own customer records, and a validator-level attack cannot expose end users at all, the architecture pushes the sharpest privacy risk down to individual institutions rather than the shared rails.

That has consequences for how banks approach custody, key management and data handling if a retail CBDC ever ships. It also frames the gap between what the prototype proves and what BIS says the stack still lacks before deployment.

For now, the message from Basel is measured: the privacy properties are achievable in a working prototype, but the engineering, institutional safeguards and breach-resilience still fall short of production-grade. BIS has not announced a date for a third phase, but the Phase 2 report establishes the technical baseline any future privacy-preserving CBDC design will be measured against.

By Daniel Okafor