On February 17, 2026, StarkWare announced the integration of EY's Nightfall zero-knowledge privacy layer into Starknet — giving institutions the ability to run confidential payments, treasury operations, and DeFi strategies on public Ethereum rails for the first time. The timing is not accidental...
"Privacy is the missing ingredient for large-scale enterprise payments on-chain. The only scalable way to achieve it is on a public, shared blockchain built on open standards." — Paul Brody, Global Blockchain Leader, EY
On February 17, 2026, StarkWare announced the integration of EY's Nightfall zero-knowledge privacy layer into Starknet — giving institutions the ability to run confidential payments, treasury operations, and DeFi strategies on public Ethereum rails for the first time. The timing is not accidental. Within the same week, Google Cloud confirmed it will operate a validator node for the Midnight Network ahead of its March mainnet launch, and Chainlink's Confidential Compute service entered early access through its Runtime Environment.
Three distinct privacy architectures are now racing to become the default institutional layer for on-chain finance. The stakes are enormous: stablecoin and tokenized payment volumes surpassed $10 trillion in adjusted settlement volume in 2025, yet the vast majority of enterprise treasury flows remain off-chain precisely because public blockchains expose every transaction to the world. The protocol that solves privacy-with-compliance first will capture the plumbing of institutional crypto for the next decade.
This report maps the competitive landscape, evaluates the technical trade-offs, and assesses which approach is best positioned to win the $50 billion institutional Layer 2 market projected by year-end.
The institutional adoption thesis has always had a paradox at its center: the same radical transparency that makes public blockchains trustworthy also makes them unusable for enterprise finance. A Fortune 500 treasurer cannot route a $200 million supplier payment through Ethereum if every competitor, counterparty, and regulator can see the amount, timing, and destination in real time.
This is not a theoretical concern. According to ZKsync CEO Alex Gluchowski, "Enterprise crypto adoption was blocked not only by regulatory uncertainty, but by missing infrastructure. Systems could not protect sensitive data, guarantee performance under peak load, or operate within real governance and compliance constraints."
The result: while on-chain stablecoin volumes have exploded past $10 trillion, institutional treasury management remains overwhelmingly off-chain. Banks and asset managers have tokenized bonds, launched stablecoin pilots, and built custody infrastructure — but the actual movement of enterprise capital on public rails has been limited to controlled experiments.
That is now changing. Three competing privacy architectures went live or entered late-stage deployment within the same February 2026 window, each backed by institutional-grade partners and each proposing a fundamentally different answer to the same question: how do you make a public blockchain private enough for JPMorgan but auditable enough for the SEC?
The privacy infrastructure war breaks down along architectural lines:
| Feature | Nightfall (EY/Starknet) | Midnight Network | Chainlink Confidential Compute | |---|---|---|---| | Architecture | Layer 3 ZK rollup on Starknet (L2) | Standalone L1 with ZK proofs | Middleware / off-chain TEE compute | | Privacy Model | Private by default, selective disclosure | "Rational privacy" — programmable disclosure | Encrypted cross-chain transactions | | Settlement | Ethereum via Starknet | Own chain (Cardano-adjacent) | Multi-chain via CCIP | | Institutional Sponsor | EY (Big Four) | Google Cloud (validator) | ANZ, ADDX, Citi, Mastercard | | Status | Live (Feb 17, 2026) | Mainnet targeting March 2026 | Early Access (CRE, early 2026) | | Open Source | Yes (public domain) | Partial | No |
Each represents a bet on where the institutional privacy layer should sit in the stack — and each carries distinct economic implications for the protocols beneath it.
The Starknet-Nightfall integration is the most architecturally aggressive play. Nightfall deploys as a Layer 3 — an appchain-style privacy environment that sits on top of Starknet's Layer 2, which itself settles to Ethereum. The result is a three-layer stack: Ethereum for security, Starknet for scalability and ZK verification, and Nightfall for institutional privacy.
"Blockchains can offer every institution the equivalent of a private superhighway for stablecoins and tokenized deposits," said Eli Ben-Sasson, StarkWare co-founder and CEO. "With Nightfall, Starknet will show how that future actually works."
The technical architecture delivers several institutional requirements simultaneously:
The strategic insight is EY's positioning. As a Big Four accounting firm that already audits many of the target institutional clients, EY brings regulatory credibility that no crypto-native team can replicate. Nightfall is open-source and in the public domain — EY neither owns nor controls it — but the association provides what StarkWare's Alex Gruell called "regulatory fluency" that accelerates enterprise procurement cycles.
The deployment is phased: private payments and transfers first, expanding with verifier upgrades to cover lending, swaps, and yield strategies. Starknet's current TVL sits at approximately $280 million — modest by L2 standards — but the Nightfall integration is designed to attract an entirely new class of capital that does not yet exist on-chain.
Critically, Nightfall is not exclusive to Starknet. The protocol also deployed as a Layer 3 on Plume Network in September 2025, targeting institutional RWA tokenization. Plume has since secured an Abu Dhabi Global Market (ADGM) license and partnered with Securitize — the BlackRock-backed tokenization platform — with a target of $100 million in institutional capital. The multi-chain deployment strategy positions Nightfall as a vendor-neutral privacy standard rather than a single-chain feature.
Midnight takes the opposite approach: rather than layering privacy onto existing infrastructure, it builds a dedicated blockchain from scratch. The network describes its model as "rational privacy" — a framework where users and applications control what data is shared, with whom, and when, using zero-knowledge cryptography.
The mainnet launch, targeting late March 2026, comes with heavyweight infrastructure backing. Google Cloud confirmed it will operate a validator node and provide critical infrastructure through its Mandiant division for advanced threat monitoring and incident response. The network's rollout follows a phased decentralization model where a limited group of trusted federated node operators — led by Google Cloud — will manage block production during the initial launch period.
This approach has drawn criticism. Cysic founder Leo Fan publicly challenged Charles Hoskinson over the Google Cloud partnership, questioning whether relying on a hyperscaler for block production contradicts decentralization principles. The debate highlights a fundamental tension in institutional privacy: enterprises want the security guarantees of Big Tech infrastructure, but crypto-native communities view such dependencies as antithetical to blockchain's purpose.
Midnight's NIGHT token distribution established one of the broadest initial holder bases in blockchain history, and the four-phase roadmap through 2026 outlines a path from token launch to full cross-chain privacy capabilities. But the standalone-chain approach faces a cold-start problem: unlike Nightfall, which inherits Starknet's existing DeFi ecosystem and Ethereum's security, Midnight must build liquidity, developer tooling, and institutional integrations from zero.
Chainlink's entry reflects its broader strategy of becoming the connective tissue of institutional crypto without competing directly with any single chain. Confidential Compute, launching through the Chainlink Runtime Environment in early 2026, enables private smart contracts that connect to real-world financial data and Web2 systems while keeping proprietary data, business logic, and computation fully confidential.
The key differentiation is cross-chain interoperability. The Blockchain Privacy Manager enables CCIP Private Transactions across both public and private chains, keeping sensitive data encrypted and invisible to node operators and third parties. For institutions operating across multiple blockchain environments — the norm for any large financial institution — this horizontal privacy layer avoids chain lock-in.
Real-world traction is already visible. ANZ, ADDX, and Chainlink have collaborated on cross-border tokenized asset transactions using CCIP Private Transactions. Chainlink reports meetings with over 100 major financial institutions across Dubai, Abu Dhabi, Hong Kong, Singapore, and Sydney. The pipeline includes delivery-versus-payment (DvP) settlement where tokenized assets and stablecoins exist on different chains — a use case that neither Nightfall nor Midnight can natively serve without Chainlink's cross-chain infrastructure.
General availability of Confidential Compute is planned for year-end 2026, with privacy-preserving tokenization of bonds, private credit pools, and fund allocations as target use cases.
The timing of this privacy infrastructure buildout is not coincidental. MiCA's full compliance deadline of July 2026 is forcing European institutions to formalize their on-chain strategies, and over 50% of European banks are planning MiCA partnerships by 2026. Meanwhile, 78% of institutional investors view MiCA as a positive development for adoption.
But MiCA creates a specific privacy paradox: it demands KYC/AML compliance and auditability while institutions simultaneously need transaction confidentiality. The privacy architectures competing today are designed precisely to thread this needle — enabling regulatory oversight without public exposure.
ZKsync's roadmap crystallizes the institutional requirements into four standards: privacy by default, deterministic control, verifiable risk management, and native connectivity to global markets. With collaborations spanning over 30 major global institutions including Citi and Mastercard, plus two unnamed central banks, ZKsync projects that multiple regulated financial institutions will launch production systems serving "tens of millions rather than thousands" of end users.
The economic implications of this privacy infrastructure war extend beyond the competing protocols themselves. The winner will determine which settlement layer captures institutional capital flows:
Forecasts project total value locked by enterprises on Layer 2 networks to surpass $50 billion by 2026, though 70% of firms are reportedly delaying large-scale Layer 2 integration pending standardization. The privacy infrastructure race may be the catalyst that breaks this inertia — or it may fragment institutional capital across incompatible privacy standards.
Three competing privacy architectures went live or entered late-stage deployment in February 2026, each backed by institutional-grade partners: EY (Nightfall/Starknet), Google Cloud (Midnight), and Chainlink (Confidential Compute).
The $10 trillion stablecoin settlement market remains largely untapped by institutional treasuries because public blockchains lack native privacy. The protocol that solves privacy-with-compliance captures the plumbing of enterprise crypto.
Nightfall's Big Four credibility is its moat: EY's audit relationships with target institutional clients accelerate procurement in a way crypto-native teams cannot replicate. Its multi-chain deployment (Starknet + Plume) positions it as a vendor-neutral standard.
Midnight's standalone approach offers clean-sheet design but faces a cold-start problem. Google Cloud's validator role provides infrastructure credibility but raises decentralization concerns.
Chainlink's cross-chain privacy is the only solution that natively serves multi-chain institutional environments, but its middleware positioning means it depends on other chains for settlement economics.
MiCA's July 2026 deadline is the forcing function: European institutions must formalize on-chain strategies, and privacy infrastructure is the prerequisite for institutional-scale deployment.
The economic stakes are existential for Layer 2 protocols: the privacy standard that wins determines which settlement layer captures the projected $50 billion in enterprise TVL.
The blockchain privacy infrastructure war is not about privacy for its own sake — it is about who captures the economic value of institutional settlement. For a decade, enterprises have cited transparency as the barrier to on-chain adoption. That excuse is now expiring. Three distinct architectures, each backed by institutional-grade partners, are live or weeks from launch.
The most likely outcome is not winner-take-all but specialization: Nightfall for Ethereum-native institutional DeFi, Chainlink for cross-chain settlement privacy, and Midnight for standalone enterprise applications that require clean regulatory boundaries. But the window for establishing standards is narrow. As MiCA's July deadline approaches and institutional procurement cycles accelerate, the protocols that ship production-ready privacy infrastructure in Q1 2026 will define the institutional stack for years to come.
What is clear is that the privacy layer — not the execution layer, not the consensus layer — has become the decisive battlefield for institutional crypto adoption. The $10 trillion in on-chain settlement volume is waiting for permission to move through private rails. The race to build those rails is now fully underway.