Tiger Research Analyzes Maru's Vision for Korean On-Chain Financial Infrastructure

Maru, aiming to integrate Korean won payments and domestic financial regulations directly into blockchain, has presented a new design direction for Korean-style on-chain financial infrastructure. Tiger Research's report analyzes Maru's initiative to combine KRW stablecoin-based transactions, transaction fees, regulatory compliance, privacy protection, and AI agent authority management within a single network.

The project is led by Hashed Open Finance under Hashed, targeting stablecoins, real-world asset tokenization, and tokenized securities. Shard Lab, with experience in Southeast Asian payment infrastructure, and Dilight Labs, which has operated nodes and mainnets since 2018, are also participating. The three organizations maintain a physical team in Seoul to verify regulatory compliance with authorities and financial institutions while expanding use cases with academia and startups.

Maru's base payment unit is the KRW stablecoin OKRW, aiming to connect domestic economic flows—such as salary payments and tax and living expenses—in an on-chain manner. Transaction fees will also be processed in KRW-based assets to reduce the burden of considering both network congestion and cryptocurrency price fluctuations for fee payments.

Issuance rights are designed separately from general transaction execution rights, with issuance rules managed at governance and protocol levels. The report notes that implementation uses a precompiled method embedded within the chain, applying ERC-20 standards for circulation while requiring issuance through a separate system account to ensure compatibility and control.

For emergency response, Maru is considering limited state corrections under legal grounds rather than rolling back the entire chain in cases of hacking or illegal fund movement. Asset freezing, recovery, and reissuance for victim compensation are discussed, though specific activation conditions and execution procedures remain under discussion.

Tiger Research highlights Maru's approach to moving fragmented transaction rules across individual services onto a common infrastructure, focusing on 'legal oracles' and the programmable regulatory compliance layer (PCL). Legal oracles supply baseline data like transaction limits, while PCL applies these as verification conditions for actual transactions. A committee involving regulators, financial institutions, and legal bodies will manage changes through consensus and maintain change records.

The regulatory compliance layer is designed to combine conditions such as transfer blocking, authentication requirements, and time-based limits. For example, a baseline amount and 24-hour reset cycle can be used to manage transaction limits, with violations blocked before execution and reasons returned in machine-readable code.

Transaction routes are divided based on required verification levels: 'open paths' apply network-wide common policies, while 'regulated paths' combine these with additional service-specific authentication and limit conditions. Financial institutions may consider transactions through regulated paths as settlement targets based on their own criteria.

For financial information protection, Maru proposes shielded pools using zero-knowledge proofs and role-based access control to verify transaction validity while selectively hiding details like amounts or recipients. Institutional access is limited to approved ranges upon meeting legal and governance requirements, though this model remains at the concept validation (PoC) stage.

For AI agent transactions, Maru introduces an Agent Verification (KYA) system linking agent addresses, certified owners, delegation permissions, and spending limits in a registry. Similar to corporate cards specifying users, spending limits, and responsible parties, it references ERC-8004 for identity standards but distinguishes between standardization of identity information and actual transaction approval criteria.

Control mechanisms include separating funds deposited with agents, per-transaction spending limits, allowed recipient addresses, and consolidated limits at the owner level. For instance, a 1 million won delegation limit would block single transactions exceeding that amount, while splitting transactions across multiple agents would still apply the overall owner limit. A 24-hour restriction on transfers and policy changes after ownership transfer requests is also proposed.

Tiger Research emphasizes distinguishing between policies to be determined by regulation and technical structures that can currently be validated. Maru's future evaluation will focus not just on design scope but on whether regulatory compliance, privacy protection, and transaction control function together in real financial environments.