Hi everyone,
It has been a while since we shared a broader update with the DAO. We wanted to share what the team has been working on, how our thinking has evolved, and where we see Scroll heading next.
The short version is that we have been using AI extensively inside Scroll, building a connected set of AI products, and rethinking what the Scroll network should be optimized for. We recently started sharing this new direction publicly, and we wanted to bring the conversation to the forum before coming back with a more formal DAO proposal.
Background: Using AI Inside Scroll
Over the past few months, we have made AI a priority across both our operations and product development. One of the clearest results has been in core chain operations. Work that previously required more than 20 engineers is now run by a much smaller number of dedicated engineers supported by AI agents, while maintaining our established uptime and security standards.
Using AI this way also made one thing clearer. The point is not to put the largest model on every task. Different jobs need different models, depending on performance, privacy, reliability, and cost.
The harder lesson came from running critical infrastructure.
Agents can automate a significant amount of work, but they should not have unrestricted control over the systems they operate. If an agent can act, someone still has to define what it is allowed to do, what it cannot do, and who can change those rules. Those permissions and boundaries need to stay verifiable, and they need to stay under the owner’s control.
This is the gap blockchain and ZK can fill. Blockchain can provide a trusted coordination layer for defining and enforcing rules, while ZK can help agents interact with credentials, private data, and compute without unnecessarily exposing the underlying information.
Scroll has spent years building ZK infrastructure and operating a production network. We believe those capabilities can now support a different problem: building secure and private infrastructure for AI agents.
That is what led us to Compass, and to the broader product stack we are now building.
The Products We’re Building
What may have looked like several separate projects is increasingly becoming one connected stack:
- Compass is the iOS app that will allow users to access the LLM models and features that come with Compass API, alongside VPN and eSIM functionality.
- Compass API is the AI routing infrastructure app that provides access to 30+ LLM models.
- CENO provides the trust layer, securing credentials, agent context, and privacy-preserving compute interactions through ZK technology.
- USX is intended to support payments and settlement across the product stack.
- SCR will continue to coordinate governance and may support broader ecosystem participation as the products develop.
- Scroll Network provides the trust and coordination layer, ensuring that agent owners remain in control of both rules and data.
Compass is where demand comes in. Compass API routes it. CENO protects the interaction. USX supports settlement. SCR coordinates governance. The network is what holds the rules and the trust in place.
What We’re Seeing So Far
These products are still at different stages.
CENO has reached more than 30 prospective customers, with 12 currently participating in proof-of-concept (POC) testing. This potential pipeline gives us an early indication of where enterprise demand may be.
Compass remains in early consumer validation, and we are continuing to build Compass API and orient the app around that product direction.
We have also been rebuilding our content and distribution capabilities across Instagram, YouTube, TikTok, X, and LinkedIn. One of our Instagram pages has grown to more than 17,000 followers, and our strongest posts have generated more than five million views. The next step is turning that reach into active users, repeat usage, and revenue.
Where Scroll Is Heading
All of this work has led us to reconsider the role of the network itself.
We no longer believe that remaining a general-purpose public chain is the best way for Scroll to differentiate or support the products we are building.
As AI agents become more capable, they will increasingly interact with credentials, proprietary data, payments, private workflows, and compute resources. Those interactions require more than execution. They require clear rules, mechanisms for verifying that those rules are followed, and ways to protect sensitive information.
For that reason, we are planning to evolve Scroll from a general-purpose public network into a private, application-specific network designed around Compass and our broader AI product stack.
We recently began sharing this direction publicly:
There is still a lot to work through, including the network’s exact operating model and the detailed transition path. Our current expectation is that the broader transition will happen gradually over nine months.
As part of the current plan, SCR would move to Ethereum mainnet so it can remain broadly accessible and liquid. There will be no change to SCR’s supply or overall tokenomics, and SCR will continue to serve as the governance token for the Scroll network.
What Comes Next
This post is simply an update, not a formal proposal or vote.
We wanted delegates and the community to see how the pieces fit together before the formal DAO process begins.
We will come back with a separate proposal on the network transition, and on the decisions that need DAO approval, once those details are ready.
In the mean time, questions, feedback, and concerns are welcome below.