Scroll DAO Proposes a Pivot to a Privacy AI Network and a Move Home

November 5, 2026 MM Crypto Team

A major strategic proposal has landed in the Scroll community, and if it passes it would reshape what Scroll is for. On November 4, a proposal posted to the Scroll DAO forum proposed transforming Scroll from a general-purpose zero-knowledge Ethereum Virtual Machine Layer 2 network into a permissioned, dedicated network built for privacy-focused artificial intelligence applications. Alongside that technical pivot, the proposal would migrate the SCR governance token to the Ethereum mainnet. It is a plan to narrow Scroll’s focus, deepen its specialisation, and bring its governance token home to Ethereum itself.

The proposal matters because Scroll has spent more than three years building an identity as a general-purpose Layer 2. Its mainnet launched in October 2023, and distribution of its SCR token was completed in October 2024. For 37 months, Scroll has operated as a general-purpose Layer 2, open to a wide range of applications. The new proposal argues, in effect, that breadth is no longer the best strategy, and that a dedicated focus on privacy AI offers a clearer path forward. That is a significant change of direction, put to the community through DAO governance rather than imposed from above.

It is important to be clear about status. The proposal is currently in the forum discussion phase, when community members can read it, debate it, raise objections and suggest changes before any formal vote. According to the process described, the next steps would be thermometer checks followed by a formal governance vote. Until that vote happens, nothing is decided. The proposal is a serious, detailed plan, but it is not yet a decision, and readers should not treat it as though Scroll has already changed direction.

From General-Purpose Layer 2 to a Dedicated Privacy Network

As a general-purpose zkEVM Layer 2, Scroll was designed to offer Ethereum-compatible smart contract execution with the scalability benefits of a Layer 2, using zero-knowledge technology to secure transactions. Developers could build many kinds of applications on it, from payments to games to decentralised finance, with lower costs and faster confirmation than on Ethereum mainnet. That openness is the standard general-purpose model, and its advantage is that anyone can build and the network does not pick winners in advance.

The proposal would move away from that model toward a permissioned network dedicated to privacy AI applications. Permissioned here means participation is governed by rules about who can do what, rather than being entirely open to all comers. That may sound like a step away from permissionless ideals, but the proposal frames it as a practical requirement: AI applications handling sensitive data, managing budgets and paying providers need clear rules about agent permissions, spending limits and settlement. A network enforcing those rules on-chain can give users and developers stronger guarantees than one leaving them entirely to individual applications. It also reflects a broader Layer 2 debate, as growing competition among general-purpose networks makes differentiation harder, while a specialised network could offer tools general-purpose rivals do not.

What Happens to SCR and the Move to Ethereum Mainnet

One of the most closely watched parts of the proposal is what happens to SCR, Scroll’s governance token. Under the plan, SCR would migrate to the Ethereum mainnet, and its supply would remain unchanged at 1 billion tokens. The migration is described as non-dilutive, with allocation unchanged, meaning existing holders would not see their proportional ownership reduced by new tokens. For holders, those details matter enormously, because a migration changing supply or allocation would be a very different proposition from one preserving both.

SCR would also keep its governance role, extended to cover USX core parameters. USX is central to the new design, and giving SCR holders a say over its core parameters ties governance of the token to governance of the stablecoin powering the new network’s economy. The move to Ethereum mainnet is significant in its own right: Ethereum mainnet is the most secure and liquid environment in the ecosystem, and hosting a governance token there can improve accessibility and integration with wider decentralised finance. Symbolically, it also signals a closer relationship with the base layer. Scroll began as a Layer 2 built to extend Ethereum, and bringing its governance token back to Ethereum mainnet is, as the proposal frames it, a move home.

USX, Compass, CENO and the Privacy AI Product Stack

The proposal describes a product stack for privacy AI, naming Compass, the Compass API, CENO, USX and SCR itself. Together they are intended to cover the journey from building a privacy AI application to running it, paying for it and settling with providers. Compass and its API appear to be the developer-facing tools, while CENO sits alongside them in the stack. USX and SCR provide the economic layer: USX for payments and budgeting, SCR for governance.

USX deserves particular attention. It is described as a dollar-pegged stablecoin for AI service pricing, task budgeting and provider settlement. In practical terms, USX would be the unit in which AI services are priced, the unit in which a user or agent sets a budget for a task, and the unit in which providers are paid when work is completed. Stablecoins suit that role because their dollar peg makes pricing predictable, unlike a volatile token. For AI tasks, where costs can accumulate across many small steps, predictable pricing and clear budgeting are essential, not extras. Privacy is delivered through zero-knowledge proofs, which let one party prove something is true without revealing underlying data, plus on-chain rules for agent permissions, spending limits and settlement. Together they aim to solve a real problem: how to let autonomous agents act for a user without giving them unlimited access to data or funds.

Why the Scroll DAO Privacy AI Pivot, and Why Now

The timing reflects a growing recognition that artificial intelligence and crypto are converging, and privacy is one of the hardest problems at that intersection. AI systems increasingly work with personal and commercially sensitive data, while AI agents become more autonomous, taking actions, spending budgets and interacting with services on behalf of users. Those trends create demand for infrastructure that can prove what happened, limit what an agent can do, and settle payments transparently, all without exposing underlying data to everyone.

Scroll’s history makes this pivot both plausible and risky. On one hand, the project has deep expertise in zero-knowledge technology, directly relevant to privacy-preserving computation, and a team that spent years building a zkEVM Layer 2 is well placed to build a network using zero-knowledge proofs for AI workloads. On the other hand, pivoting away from a general-purpose model means leaving behind the broad developer base built over 37 months, and some existing applications may not fit the new dedicated model. The community debate will need to grapple with what happens to them. The proposal notes technical design is complete, with implementation planned for the end of 2026 and a fuller timeline to be announced. That sequencing allows discussion and refinement before resources are committed, but it also means readers should not expect an overnight transformation.

What the Community Decides Next

The immediate future lies in the Scroll DAO forum, where discussion is now under way. Forum discussion is the first phase of decentralised governance: it surfaces concerns, tests arguments, and lets authors refine the plan in response to feedback. Thermometer checks, which gauge sentiment informally, would follow, and only then would a formal governance vote decide the matter. That staged process is designed to prevent rushed decisions on a change of this magnitude.

For SCR holders, key questions are likely to concern migration mechanics, governance of USX core parameters, and treatment of existing applications. For developers, questions are likely to concern the permissioned model: who can build, what permissions are required, and how the new tools compare with the general-purpose environment they know. For outside observers, the proposal is a case study in Layer 2 evolution. The early Layer 2 story was about scaling Ethereum for everyone; the next chapter may be about specialisation, with different networks serving different niches. Scroll’s proposal, if adopted, would be one of the clearest examples of that shift, and whatever the outcome, it shows a major Layer 2 willing to reconsider its strategy in public through DAO governance.

FAQ

What is the Scroll DAO proposing?

The November 4 Scroll DAO forum proposal proposes transforming Scroll from a general-purpose zkEVM Layer 2 into a permissioned, dedicated network for privacy AI applications, and migrating the SCR governance token to Ethereum mainnet. It is in forum discussion, with thermometer checks and a formal governance vote to follow.

Will the SCR supply change under the proposal?

No. SCR supply would remain unchanged at 1 billion tokens, the migration would be non-dilutive, and allocation would be unchanged. SCR would keep its governance role and extend it to USX core parameters.

What is USX?

USX is a dollar-pegged stablecoin described in the proposal for AI service pricing, task budgeting and provider settlement. It would be the unit in which AI services are priced, task budgets are set, and providers are paid, giving predictable pricing for AI workloads.

What products are part of the new privacy AI stack?

The proposal names Compass, the Compass API, CENO, USX and SCR. Together they support building, running and settling privacy AI applications, with privacy delivered through zero-knowledge proofs and on-chain rules for agent permissions, spending limits and settlement.

When would the pivot be implemented?

Technical design is complete and implementation is planned for the end of 2026, with a fuller timeline to be announced. Before implementation, the proposal must pass forum discussion, thermometer checks and a formal governance vote, so timing depends on that governance process.

Source: KuCoin/Cointime

MM Crypto Analysis Web