Akita Moves Post-Quantum Security Into the Polynomial Commitment Layer

LayerZero announced Akita on September 9 as a lattice-based polynomial commitment scheme designed to add post-quantum security to zero-knowledge proving systems. The commitment layer is where a prover binds itself to computation data before opening selected evaluations for a verifier, so changing it can affect proof size, proving cost, memory use, setup assumptions, and compatibility across the stack.

Akita’s repository describes the scheme as modular, post-quantum, and based on a transparent setup. That combination matters for builders because it targets two operational concerns at once: reducing dependence on cryptographic assumptions vulnerable to quantum attacks and avoiding a trusted setup ceremony for the commitment component.

LayerZero says Akita produces 65–80 KB proofs, compared with 200 KB or more for the hash-based post-quantum systems it compares against. It also claims two- to three-times faster proving than the current Jolt implementation and roughly half the memory use. Those figures are announcement claims, not an independent benchmark; teams should reproduce them with their own circuits, hardware, field choices, and verifier environment before treating them as deployment forecasts.

The first announced integration is Jolt, the zkVM associated with the Zero blockchain. The practical angle is architectural rather than cosmetic: Akita is presented as a standalone primitive, while Jolt’s codebase already contains a dedicated jolt-akita workspace member and pins Akita crates from the LayerZero repository. That makes the commitment scheme an explicit dependency boundary that can be reviewed, versioned, and tested.

There is an important status distinction. LayerZero describes the Jolt deployment as the first intended deployment, while the current JoltBook still says Jolt uses Dory as its polynomial commitment scheme. The evidence supports “integration announced and under development,” not “Akita is already Jolt’s production default.”

For ICP and other zkVM builders, the takeaway is to audit the commitment interface as a first-class migration surface. A post-quantum claim is not enough: verify transcript binding, serialization formats, proof-version compatibility, setup assumptions, verifier implementation, and reproducible benchmarks. Akita makes that checklist more visible by putting the post-quantum transition in a modular PCS rather than framing it as a wholesale zkVM replacement.