Interchain Security has undergone multiple iterations since the concept first emerged. In a 2018 talk on interchain scaling and security models, Jae Kwon laid out different ways that chains could relate to each other from a security perspective. One of those models was what he called replicated shared security.
Instead of every chain standing up its own validator set, the same set could be reused across multiple chains, allowing them to inherit the same economic security. Implicit in that model was a shift in how validators were treated. They were no longer just block producers for a single chain, but security providers operating across chains. Chain security could become something that could be offered, consumed and paid for explicitly. In other words, “Validation As A Service” or “VaaS.”
As AtomOne prepares to support consumer chains of its own, it’s worth looking at how those previous versions translated into real systems. Today, only two chains still run under Cosmos’ Interchain Security (ICS) even though the intention was for there to be dozens. Why did adoption stall?
The ICS protocol looked fine on paper, but in practice it treated validator labor as elastic. The economics could not sustain the validators expected to run it. Even with later improvements (Partial Set Security), ICS remains unsustainable for most validators. Sometimes the issue is infrastructure costs, other times it’s the time and attention required to maintain consumer chains. These are the problems VaaS will focus on.
Validation as a Service
VaaS is based on Replicated Security. All validators must validate the declared consumer chains. VaaS is a revamp of the latest interchain-security model, with features intentionally removed where they added complexity without clear operational benefit. A return to the original vision, stripped down to what actually matters. A technical overview is available in the README.
The biggest change is economic. Instead of relying on ICS rewards, VaaS replaces them with a fixed fee in PHOTON payable for each block. At every block, a parameterized amount of PHOTON is distributed evenly to validators. If that balance runs out, the provider chain instructs the consumer chain to halt.
In this new model, validators are not expected to run consumer chains at a loss, and consumer chains know exactly what they are paying for security and when that security stops. This aims to be better than ICS rewards because a single token is used (always PHOTON) and is not tied to inflation. This is designed to make validator compensation more predictable, which was not the case before. We expect this to reduce situations where validators absorb uncompensated operating costs.
Looking Ahead
Interchain Security showed that shared security models in Cosmos can work on a technical level. It also showed that security models collapse when incentives and operations do not match reality. Validation as a Service builds directly on those lessons. By simplifying assumptions and treating validation as an explicit service, it makes shared security easier to sustain.
As AtomOne moves forward, VaaS is one part of a broader effort to make the network more modular and practical to operate. Governance upgrades, DAO activations and continued work on validation improvements are all moving in parallel.
VaaS is under active development and aimed for a v5 release. We are targeting Gno.land as the first VaaS consumer chain and plan to bring it online soon, pending final readiness milestones. Ready yourselves!