# Privacy Layer-2s Ship Rival "Guardian" Designs, Turning the Off-Switch for Confidentiality Into a Contest

Aztec, Miden, and an Ethereum research proposal each answer the same question differently: who holds the key that can lift privacy, the user, a committee, or the state.

- Published: 2026-07-17T05:28:14.121Z
- Canonical: https://polylog.news/crypto/2026-07-17/privacy-layer-2s-ship-rival-guardian-designs-turning-the-off
- Publisher: Polylog (Crypto desk)
- Section: crypto
- Sources: [Ethereum Research](https://ethresear.ch/t/privacy-guardians-2-0/25471), [Aztec Network](https://aztec.network/blog/who-controls-your-privacy-off-switch), [Miden](https://miden.xyz/blog/what-is-miden-guardian)

Within a single day, three teams building privacy into public blockchains published designs that converge on one unresolved problem. A confidential system needs a way to reveal information under some conditions, and whoever controls that reveal controls the privacy.

An Ethereum researcher writing under the name Leo Glisic [open-sourced a design called Privacy Guardians 2.0](https://ethresear.ch/t/privacy-guardians-2-0/25471), a scheme for selective disclosure that lets a user prove specific facts about a shielded transaction without exposing the whole record. Aztec, a zero-knowledge (a method of proving a statement true without revealing the underlying data) rollup, framed the same tension directly in a post titled [who controls your privacy off-switch](https://aztec.network/blog/who-controls-your-privacy-off-switch), arguing that a compliance access path held by an operator is not privacy at all. Miden, a separate privacy chain, described [Miden Guardian](https://miden.xyz/blog/what-is-miden-guardian) as client-side infrastructure that keeps the disclosure key with the account holder rather than a central party.

What the three share is a rejection of provider discretion. Each design tries to move the decision to reveal away from a company that can be subpoenaed or breached and toward cryptography that the user controls. That distinction is the entire argument. A privacy chain with an operator-held master key becomes, in a crisis, an ordinary database that can be surveilled.

Under a sound-money and first-principles read, this is the privacy sector maturing from marketing into architecture. The live question is no longer whether these networks hide balances, but whether that concealment survives a determined state or a hostile validator set. The three published designs are competing answers, and builders will decide which one survives in practice.

## What this means

Confidential execution is being reframed as a prerequisite for institutional on-chain use, and the value now accrues to whoever designs a disclosure mechanism regulators accept without handing an operator a master key. Users of custodial or operator-gated privacy tools are exposed to the same subpoena and breach risk as any centralized database, while networks that push the disclosure key to the client shift that liability onto the account holder. The channel is trust: cryptographic enforcement versus a company's promise.

## What to watch

- Whether any regulator or auditor formally accepts a client-side selective-disclosure proof as compliant, which would validate the entire approach against a supervised alternative.
- Adoption metrics on Aztec and Miden mainnets, since a design that no one deploys settles nothing about who controls confidentiality in practice.
- Whether a rival design emerges with an operator-held key marketed as "compliant privacy," which would signal the surveillance-friendly branch is commercially preferred.