The Coordination Layer Is Catching Up: DAO Governance and zk-Identity in 2026

This week, five developments in roughly 96 hours confirmed that Open Money's coordination layer — governance tooling, onchain identity, protocol decision-making — is entering the same structural inflection the settlement layer went through in 2024.

• Celo hit 840,000 daily active users as a mobile-first L2.
• Canton Network connected permissioned institutional rails to 165+ public chains via LayerZero.
• Tally, the dominant DAO governance platform, announced it is winding down.
• ZKredit launched the first production zk-identity primitive for undercollateralized DeFi lending.
• Balancer Labs shut down its corporate entity and transitioned fully to onchain DAO control.

Each event is significant on its own. Together, they mark the week the coordination layer stopped being the lagging indicator of open money and started being the active front.

What Is the Open Money coordination layer?

The Open Money Framework evaluates financial infrastructure across three layers: settlement rails (how value moves), intermediation services (DeFi lending, trading, yield), and the coordination layer — the governance systems, onchain identity primitives, and decision-making infrastructure that sits above both.

Permissionless settlement rails and composable services cannot self-sustain without a coordination layer that is equally open. Governance that depends on off-chain companies or centralized identity systems introduces the same capture risk that permissionless rails were designed to remove.

This is the layer that has consistently lagged. Through 2024 and into 2025, settlement and intermediation matured measurably: L2 transaction costs fell, stablecoin volumes crossed institutional thresholds, real-world asset tokenization moved from pilot to production.

Coordination remained dependent on off-chain corporate wrappers — venture-backed governance platforms, hybrid DAO-plus-company structures, and identity systems that required users to expose data to a third party to prove anything about themselves. This week changed the picture on all three fronts simultaneously.

The rails case is now settled

Celo's 840,000 daily active users prove permissionless scale is real

On March 26–27, Celo marked its one-year anniversary as an Ethereum L2 with a reported 840,000 daily active users, placing it among the most active L2s by any activity measure.

The metric matters for what it required: no identity verification, no institutional onboarding infrastructure, no web-browser-centric product layer. Celo's interface is a mobile application. The users transacting through it are doing so on the same settlement layer as institutional DeFi participants.

Permissionless access at the settlement layer is not a design property to be proven in theory anymore. It is a production outcome at consumer volume. The Permissionless Access cell of the settlement layer is filled.

Canton Network and LayerZero make institutional composability bidirectional

On March 26, Canton Network — a permissioned institutional blockchain backed by Goldman Sachs and others — announced an integration with LayerZero enabling compliant cross-chain transfers of institutional assets to 165 or more public permissionless chains. The mechanism is standardized cross-chain messaging rather than custom bridges or gatekeeper-controlled transfers.

The structural point is the direction of the choice. A closed system built explicitly for institutional compliance has elected to open outward, using the same cross-chain messaging infrastructure that permissionless DeFi protocols use. Composable Infrastructure at the settlement layer now flows in both directions — from permissionless to institutional and back.

The Coordination Layer Under Stress

Why Tally's Wind-Down Is a Structural Signal, Not a Failure

Tally, the dominant onchain governance platform serving more than 500 DAOs, announced it is winding down core operations, citing sustainability and funding challenges.The announcement has been read in some quarters as a governance failure or a market signal that DAO tooling has no business model. That reading is too narrow.

The venture-backed SaaS model for governance tooling faces a structural contradiction: protocols want decentralization, resilience, and protocol-native logic; SaaS businesses need recurring revenue, user retention, and centralized product control.

Tally was not out-competed by a better SaaS product. It was outpaced by the protocols it served, which are building governance logic directly into their own onchain systems. The Programmable Logic cell of the coordination layer is where the pressure is now concentrated.

What Tally's wind-down surfaces is an open question about what replaces centralized tooling at this layer — a question the research backlog section of this issue addresses directly.

Balancer's full DAO transition removes the last corporate wrapper

On March 24, Balancer Labs announced the shutdown of its corporate entity and a full transition to onchain DAO governance. The protocol now operates with zero token emissions, onchain fee buybacks, and pure protocol-owned governance. No company sits between users and the code.

Balancer Labs to shut down following $110 million exploit, co-founder says in DAO post

Co-founder Fernando Martinelli said he considered winding down the entire protocol but decided the team deserved a chance to restructure, with the DAO targeting zero emissions, fee restructuring, and a BAL buyback to offer holders a fair exit.

CoinDeskShaurya Malwa

Balancer's transition is the clearest concrete example this week of the Sovereign Custody dimension operating at the coordination layer. When Balancer removed the corporate wrapper, nothing broke. The hybrid model — corporate entity managing operations while a DAO nominally governs — proved unnecessary. The protocol is more composable now that the governance interface is open infrastructure rather than a company's product.

The Missing Primitive Arrives

ZKredit and the zk-Identity Problem

Settlement infrastructure can be permissionless by construction. Coordination infrastructure cannot. Governance systems need to know who has standing to vote. Lending protocols need creditworthiness signals to extend credit without demanding overcollateralization. The question the coordination layer has not been able to answer cleanly is how identity can be verified without the user surrendering control over the underlying data.

On March 25, ZKredit — built by Primus Labs and Brevis and launched on BNB Chain — announced production deployment of a zkTLS and zero-knowledge proof system for onchain credit identity.

The system allows users to prove facts about their real-world identity and credit history onchain without exposing the underlying data. A lending protocol composes against a verified proof. The user retains custody of what the proof was derived from. Neither the protocol nor any third party holds the identity data.

This addresses two cells in the coordination layer simultaneously: Transparent Verification (the proof is auditable onchain) and Sovereign Custody (the user controls the underlying data).

No prior production system at this layer has accomplished both. The existing alternatives require either exposing data to a third party or forgoing the benefits of real-world identity entirely.

What ZKredit does not yet solve

The current deployment is limited to BNB Chain [4]. Multi-chain adoption requires developer integration, security audits at each new deployment, and protocol-side changes at any lending or governance system that wants to compose against the proof.

The primitive is technically sound and live at production scale. The path from single-chain deployment to the multi-chain coordination infrastructure the flywheel requires is an execution problem, not a structural one — but it is a real problem with a real timeline measured in months, not days.

How these events connect: A coordination flywheel

The five events this week are related by the same underlying economic pressure, not by coincidence of timing.

Consumer-grade L2 scale (Celo's 840,000 DAUs) creates demand for coordination systems that can handle equivalent volume with equivalent permissionlessness. When settlement rails prove they can serve hundreds of thousands of users without gatekeepers, governance and identity infrastructure that cannot match that scale and openness becomes the binding constraint.

Institutional composability (Canton and LayerZero) creates demand for coordination primitives that can differentiate between participants without excluding them.

When institutional capital and retail permissionless activity flow through the same cross-chain layer, the identity and governance systems above them need to handle both — which is exactly what ZKredit-style verified proofs enable without forcing a data exposure trade-off.

Governance tooling consolidation (Tally's wind-down and Balancer's transition) removes friction from the coordination layer. When governance logic moves fully onchain, protocols can compose against each other's governance interfaces as open infrastructure. The coordination layer becomes more interoperable precisely because the closed corporate product layer above it is being removed.

The economic incentive structure that drove L2 scaling and institutional composability — lower costs, higher reliability, more users — is now operating identically at the coordination layer. The same forces that made closed settlement rails irrational once open rails achieved parity are making closed governance tooling irrational as open, onchain alternatives reach adequate maturity.

What could slow this down

Short-term fragmentation in governance tooling

Tally's wind-down leaves more than 500 DAOs needing new governance infrastructure. The available alternatives — native onchain governance modules, alternative platforms, protocol-specific implementations — are not uniformly mature.

Some DAOs will experience real operational disruption. Some will make choices that increase rather than decrease their off-chain dependencies during the transition. The coordination layer's evolution is not automatically smooth because the direction is clear.

Adoption lag in zk-identity

ZKredit is one implementation of zk-identity on one chain [4]. Real-world credit data verified on BNB Chain is not composable with a lending protocol on Ethereum without a separate deployment, a bridge, or a port — and each step reintroduces complexity.

Developer education, protocol integration work, and security audits at each new chain are not trivial. The approach can be technically correct while adoption remains slow for reasons that have nothing to do with the protocol's quality.

Compliance rail constraints in institutional composability

Canton's integration with public chains via LayerZero is a composability milestone within a permissioned system. The compliance rails governing which institutional assets can flow to which public chains are controlled by Canton's network participants, not by open governance.

Full bidirectional composability between institutional and permissionless systems is several steps from where this integration leaves things.

These are real friction points. They are also the kind of friction that appears at every layer of the open money stack before that layer matures. Ethereum L2s faced throughput ceilings, fragmented liquidity, and inadequate bridging infrastructure before the current solutions made those problems manageable.

The coordination layer is at an earlier point in the same maturation curve, not in a structurally different position.

Where the matrix stands

The Open Money Framework maps five dimensions — Permissionless Access, Transparent Verification, Programmable Logic, Composable Infrastructure, Sovereign Custody — across three layers: settlement rails, intermediation services, and coordination governance.

Two years ago, the settlement layer had partial scores across most dimensions. The intermediation layer was emerging. The coordination layer was underdeveloped across nearly every dimension. This week's cluster of events moves the coordination layer from underdeveloped to actively contested — the stage that historically precedes rapid maturation.

Celo fills Permissionless Access at the settlement layer definitively. Canton's LayerZero integration fills Composable Infrastructure across the settlement-to-coordination boundary.

Tally's wind-down and Balancer's transition apply pressure on Programmable Logic and Sovereign Custody at the coordination layer, forcing the evolution of more native mechanisms. ZKredit fills Transparent Verification and Sovereign Custody at the coordination layer in a way no prior production system has.

The matrix is not complete. The coordination layer still has gaps. What this week demonstrates is that the gaps are being addressed by primitives that are technically sound, economically incentivized, and deployed at production scale.

Research backlog

The following questions are open as of this issue. They are either unanswered by the week's events, underexplored in current coverage, or actively contested in ways that deserve dedicated treatment.

What replaces Tally at the tooling layer? Five hundred-plus DAOs need new governance infrastructure. The candidates — Aragon, Governor Bravo forks, custom onchain modules — have different trade-offs in terms of maturity, security, and customizability. A comparative analysis of the available options and their adoption velocity post-Tally announcement would clarify where the tooling layer is actually heading, not just where theory suggests it should.

Does Balancer's fee buyback model work as a sustainability mechanism at scale? Balancer has moved to onchain fee buybacks as its primary economic mechanism. This has not been tested through a full market cycle without a corporate entity managing treasury. The question is whether protocol-owned governance and fee-driven sustainability can maintain development velocity and respond to governance attacks without the organizational infrastructure a company provides.

When does ZKredit port to Ethereum and other major L2s? The BNB Chain deployment is a proof of concept at production scale. The timeline for multi-chain deployment — and specifically deployment to Ethereum mainnet and the major L2s where the DeFi lending volume is concentrated — will determine whether this week's launch is a primitive or a footnote. No public roadmap has been released.

What is Canton's actual compliance envelope? The LayerZero integration enables cross-chain transfers of institutional assets to 165+ public chains, but the practical question is which asset classes and under what regulatory conditions. Which jurisdictions' compliance requirements does Canton's network actually satisfy? What categories of institutional assets can flow today versus what requires additional regulatory clarity? The announcement describes a capability; the practical envelope remains underspecified.

Can onchain governance be public-goods funded rather than VC-backed? Tally's wind-down exposes the business model problem for governance tooling. The most durable alternative may not be a better SaaS company but governance tooling funded as a public good — through protocol treasuries, retroactive public goods funding mechanisms, or direct onchain grants. The precedents from Ethereum's ecosystem (Protocol Guild, Optimism's RPGF) suggest this model can work; the question is whether it can scale to the coordination needs of 500+ DAOs with diverse governance structures.

How does zk-identity interact with evolving KYC regulation? ZKredit allows users to prove creditworthiness onchain without exposing personal data. Regulators have not yet addressed whether a zero-knowledge proof of credit identity satisfies KYC requirements in lending contexts — a gap that will become relevant the moment an under-collateralized lending protocol using ZKredit faces regulatory scrutiny. The legal question may determine the adoption ceiling more than the technical question does.

Does Celo's mobile-first model generalize? Celo's 840,000 DAUs were driven by a mobile-first, KYC-free application [1]. No other top L2 has achieved comparable consumer penetration through the same approach. The question is whether Celo's strategy is repeatable — whether another L2 could build to similar scale with similar architecture — or whether Celo's position is path-dependent on specific geographic markets and payment use cases that other L2s cannot replicate.