Smart accounts decouple identity from assets. A user's intent is executed by a network of permissionless third-party actors like bundlers and paymasters, not a single custodian. This fractures the legal concept of a 'transmitter'.
account-abstraction-fixing-crypto-ux
Blog
Why Abstracted Accounts Demand a New Taxonomy of Money Transmitters
Account abstraction's core innovation—splitting user operations across bundlers, paymasters, and key managers—has inadvertently created a regulatory black hole. This analysis maps the fractured compliance landscape and argues for a new, functional taxonomy for money transmission.
introduction
THE REGULATORY MISMATCH
The Compliance Black Hole of Smart Accounts
Account abstraction's user-centric design creates a compliance vacuum where traditional money transmitter laws fail to map onto new transaction flows.
Compliance liability becomes ambiguous. Under the BSA, who is the Money Services Business: the wallet developer (Safe, Argent), the bundler infrastructure (Stackup, Pimlico), or the paymaster sponsoring gas? Current law assumes a single liable entity, not a modular stack.
Intent-based architectures obscure transaction paths. Protocols like UniswapX and CowSwap route orders through solvers, creating multi-hop settlements that defy geographic-based compliance. The 'travel rule' breaks when the originator and beneficiary are abstracted smart contracts.
Evidence: The EU's MiCA regulation explicitly carves out 'self-hosted wallets,' but a Safe multisig with a social recovery module blurs the line between hosted and non-custodial, demonstrating the taxonomy's inadequacy.
key-trends
WHY ABSTRACTED ACCOUNTS DEMAND A NEW TAXONOMY
The Fractured VASP: Three Core Trends
Smart accounts and intents are dismantling the traditional Virtual Asset Service Provider model, forcing a redefinition of what constitutes a money transmitter.
01
The Problem: The Externally Owned Account (EOA) is the Ultimate VASP
In the EOA model, the private key holder is the sole money transmitter. Every transaction is a direct, signed command. This creates a monolithic compliance target where the wallet provider (e.g., MetaMask) is legally ambiguous, but the user is fully liable.
- User is the VASP: They control the signing key, manage gas, and approve every transfer.
- Wallet as a Viewer: Frontends are just interfaces, not financial service providers.
- Regulatory Blind Spot: This model collapses under delegation, which is the core promise of account abstraction.
1
Monolithic Entity
100%
User Liability
02
The Solution: Intent-Based Architectures Fracture the VASP Function
With intents, users declare what they want, not how to do it. Execution is delegated to a network of solvers (e.g., UniswapX, CowSwap, Across). This splits the 'transmission' of value across multiple, specialized parties.
- User as Declarant: Submits a signed intent, not a transaction.
- Solver as Executor: Competes to fulfill the intent, bundling operations across chains via protocols like LayerZero.
- New Liability Stack: The VASP function is now distributed between intent originator, solver network, and settlement layer.
N
Distributed Actors
~500ms
Solver Competition
03
The New Taxonomy: Separating Custody, Intent, and Execution
Abstracted accounts (ERC-4337, Starknet, zkSync) legally separate the roles traditionally bundled in an EOA. This demands a regulatory framework that recognizes distinct entity types.
- Custodial VASP: Holds the smart account's signing authority (e.g., a social recovery module or MPC service).
- Intent Orchestrator: Aggregates and routes user intents to solver markets.
- Execution Carrier: The party (solver, sequencer) that finally submits the transaction bundle to the blockchain.
- Compliance becomes granular, applying to specific actions, not the entire user wallet.
3+
Regulatory Roles
$10B+
Protected TVL
THE NEW MONEY TRANSMITTER STACK
Functional Decomposition: Who Does What Now?
Mapping the functional responsibilities of transaction execution across traditional EOAs, smart contract wallets, and intent-based architectures.
| Functional Layer | Traditional EOA (e.g., MetaMask) | Smart Contract Wallet (e.g., Safe, Argent) | Intent-Based Abstracted Account (e.g., UniswapX, CowSwap) |
|---|---|---|---|
Transaction Construction | User manually signs raw calldata | User signs userOp for a Bundler (e.g., Stackup, Alchemy) | User signs declarative intent (e.g., 'Swap X for Y at best price') |
Gas Payment & Sponsorship | User pays gas in native token | Paymaster can sponsor gas (ERC-4337), user can pay in any token | Solver network pays gas; cost is baked into intent fulfillment |
Private Key Custody | User holds seed phrase | User holds seed phrase or social recovery guardians | User holds signing key; no on-chain seed phrase management |
Nonce Management | Sequential on-chain nonce | Parallel, off-chain userOp nonces managed by Bundler | No nonce management; intents are order-flow messages |
Fee Extraction Point | Block builder via priority fee | Bundler & Paymaster extract fees via userOp markup | Solver extracts value via MEV or explicit fee in intent fulfillment |
Regulatory Touchpoint (U.S.) | User is the money transmitter | Wallet provider may be deemed transmitter | Intent solver/fulfiller is the de facto transmitter |
Execution Guarantee | Tx succeeds or fails on-chain | UserOp can be simulated & guaranteed pre-execution | Intent is fulfilled off-chain; settlement is atomic on-chain |
Typical Latency | ~12 sec (1 Ethereum block) | ~30-60 sec (bundler queue, simulation) | < 1 sec (off-chain matching) to ~12 sec (on-chain settlement) |
deep-dive
THE TAXONOMY SHIFT
Why 'Control' is the Wrong Regulatory Lens
Abstracted accounts render the traditional 'control' test for money transmitters obsolete, demanding a new framework based on programmatic delegation.
The control test fails because abstracted accounts separate asset ownership from transaction execution. A user owns the account key but delegates signing authority to a programmatic intent solver like UniswapX or Across Protocol.
Regulatory focus must shift from 'who holds the key' to 'who defines the execution path'. The user sets the intent, but a third-party solver network controls the transaction sequence and final settlement.
This creates a new liability layer. Under old rules, a wallet provider like MetaMask is not a transmitter. But an intent-based aggregator filling that wallet's order directly intermediates value, a core MSB function.
Evidence: In Q1 2024, intent-based volumes on UniswapX and CowSwap exceeded $5B, all flowing through non-custodial yet intermediary solvers. The asset never touched the user's EOA, breaking the control paradigm.
counter-argument
THE LEGAL ARGUMENT
Steelman: "It's Just Code, Not a Service"
The core regulatory defense for abstracted accounts is that smart contract logic is not a financial service.
The core legal defense for protocols like ERC-4337 Account Abstraction is that they are permissionless, immutable code. The argument states that deploying a smart contract factory is not the same as operating a money transmitter. This distinction is foundational for projects like Safe{Wallet} and ZeroDev, which provide the infrastructure, not the service.
Regulatory precedent exists for this separation. The Howey Test evaluates an 'investment contract,' not a software library. A developer using the AA Stack is analogous to a web developer using HTTPS; the protocol is a tool, not the regulated entity. This logic underpins the legal strategy for LayerZero and Circle's CCTP.
The user's intent is sovereign. In an intent-based architecture, the user signs a declarative goal, and a decentralized network of solvers (like those on CowSwap) fulfills it. The protocol is a routing layer, not a custodian. This shifts liability from infrastructure to the end-user's signed message.
Evidence: The SEC's case against Uniswap Labs focused on the interface and token listings, not the immutable core Uniswap V3 contracts. This legal theory isolates the 'service' to the front-end, protecting the underlying protocol—a blueprint AA projects will follow.
risk-analysis
WHY ABSTRACTED ACCOUNTS DEMAND A NEW TAXONOMY
The Slippery Slope: Three Regulatory Nightmares
Smart accounts and intent-based protocols are fracturing the legal definition of a 'money transmitter,' creating three distinct regulatory black holes.
01
The Bundler Liability Trap
Bundlers are the new payment processors, but current law sees them as unlicensed transmitters. They aggregate and submit user operations, touching funds but never holding private keys. This creates a direct clash with FinCEN's 'acceptance and transmission' rule.
- Legal Risk: A single sanctioned transaction could implicate the entire bundler network.
- Precedent: The OFAC sanctioning of Tornado Cash smart contracts sets a dangerous analogue for code executors.
- Scale: Top bundlers like Stackup and Pimlico already facilitate millions of user operations monthly.
1M+
Ops/Month
0 Keys
Custody Held
02
Paymaster as Unlicensed Bank
Paymasters sponsor gas fees, effectively extending credit. Regulators will view this as a money transmission service where the asset transmitted is computational resource (gas), convertible to fiat.
- Core Issue: Sponsoring fees for a user's NFT mint is functionally a micro-loan settled in ETH, triggering lending and transmission laws.
- Compliance Black Box: Protocols like Biconomy and ZeroDev cannot KYC ephemeral smart account addresses.
- Systemic Risk: A dominant paymaster failing could strand thousands of pending user operations, a novel form of settlement failure.
Gas
Asset Class
0 KYC
For Users
03
Intent Architect as Unregistered Broker
Solvers in intent-based systems (UniswapX, CowSwap) don't just route orders; they determine execution paths for a fee. This transforms them into brokers or dealers under the 'regular business' definition of the Exchange Act.
- Regulatory Overlap: SEC may claim jurisdiction if the swapped asset is a security; CFTC if a commodity; FinCEN regardless.
- Opaque Markets: Solver competition creates a non-transparent order flow market worse than traditional Wall Street dark pools.
- Market Scale: Across Protocol and Anoma-based systems already route $100M+ in cross-chain intent volume.
$100M+
Intent Volume
3 Agencies
Jurisdictional Clash
future-outlook
THE REGULATORY GAP
Proposal: A Functional Taxonomy for Money Transmission
The rise of abstracted accounts and intents dissolves the traditional legal definition of a money transmitter, creating a critical regulatory and operational blind spot.
The legal definition of a Money Transmitter is obsolete. It assumes a single, identifiable entity controls the private key and initiates the transaction. Abstracted accounts like Safe{Wallet} and ERC-4337 smart accounts separate ownership from execution, distributing control across bundlers, paymasters, and session keys.
Intent-based architectures dissolve transactional responsibility. Protocols like UniswapX and CowSwap do not execute user-signed transactions; they solve for a user's stated outcome. The entity fulfilling the solver role transmits value but never holds a user's private key, falling outside current regulatory frameworks.
This creates a systemic risk. A bundler in an ERC-4337 stack (e.g., Stackup, Alchemy) technically transmits user funds but is legally ambiguous. This mismatch stifles institutional adoption, as compliance teams cannot map roles like paymaster or solver to existing Money Transmitter Licenses (MTLs).
Evidence: The EU's MiCA regulation explicitly exempts 'software providers,' a category that now includes critical infrastructure like Pimlico's paymaster service, which directly subsidizes and facilitates the transfer of funds for thousands of accounts.
takeaways
WHY ABSTRACTED ACCOUNTS BREAK THE OLD RULES
TL;DR: The New Compliance Stack
Smart accounts and intent-based architectures dissolve the clear lines between user, application, and infrastructure, forcing regulators to look beyond simple wallet addresses.
01
The Problem: The Vanishing Custodian
ERC-4337 Bundlers and Paymasters execute transactions on behalf of users but never hold keys. Regulators can't classify them as traditional custodians (like Coinbase), yet they have ultimate settlement power. This creates a regulatory blind spot for the entities with the most operational control.
- Key Actor: ERC-4337 Bundler
- Jurisdictional Gap: Not a custodian, not a pure software publisher
- Risk: Unlicensed money transmission at the protocol layer
0
Keys Held
100%
Tx Control
02
The Solution: Intent-Based Liability
Shift focus from who holds the key to who fulfills the intent. Projects like UniswapX, CowSwap, and Across use Solvers and Fillers as licensed intermediaries. Compliance attaches to the fulfillment layer, not the signature. This creates a clear on/off-ramp for regulatory oversight.
- Key Entity: Solver/Filler Network
- Mechanism: Licensed fillers compete to fulfill user intents
- Precedent: Existing MSB frameworks can apply to fillers
Solver
Liable Entity
User
Holds Intent
03
The Problem: Programmable Compliance is a Minefield
Smart account features like social recovery, session keys, and batched transactions can be used to programmatically obscure the chain of custody. A Safe{Wallet} with a 2/3 multisig where signers are other smart contracts creates an ownership graph, not a clear sender. Automated compliance tools (Chainalysis, TRM Labs) break down.
- Example: Nested Smart Account Recovery
- Failure: AML/KYC flags cannot resolve programmable ownership
- Impact: False positives and undetectable structuring
Graph
Ownership Model
~0%
Tool Accuracy
04
The Solution: Verifiable Credentials at the Account Level
Embed attestations (e.g., Ethereum Attestation Service, Verax) directly into the account abstraction stack. A smart account's modular compliance module can require a valid credential from a licensed verifier before permitting certain actions (high-value transfers, interacting with DeFi pools). The rule is enforced by the wallet, not the chain.
- Tech Stack: EAS, Verax, Sphere
- Enforcement: Compliance as a smart account module
- Benefit: Granular, user-held, privacy-preserving checks
On-Chain
Attestation
Modular
Enforcement
05
The Problem: Cross-Chain Intents Obscure Trails
When a user expresses an intent to "swap ETH on Arbitrum for USDC on Base," systems like LayerZero and Axelar facilitate the cross-chain settlement. The user signs once, but assets move across multiple jurisdictions and validator sets. Which chain's regulators have authority? The old model of per-chain analysis fails.
- Protocols: LayerZero, Axelar, CCIP
- Gap: Multi-jurisdictional settlement in one signature
- Risk: Liability arbitrage across chains
1
User Signature
N
Chain Hops
06
The Solution: The Funnel Model & Licensed Abstractors
Treat the entry point as the choke point. Regulate the "Abstractor"—the front-end or wallet (e.g., Safe, Biconomy) that initiates the user's intent bundle. They perform KYC and travel rule, then cryptographically attest to it. Downstream actors (bundlers, solvers) can trust this attestation. This mirrors the traditional "gateway" model but for abstracted flows.
- Entity: Front-end / Wallet as Abstractor
- Duty: Initial KYC/Travel Rule, On-Chain Attestation
- Flow: Compliance flows with the intent, not the asset
Abstractor
Licensed Gate
Attestation
Passport
ENQUIRY
Get In Touch
today.
Our experts will offer a free quote and a 30min call to discuss your project.
NDA Protected
Confidentiality24h Response
Time to ValueDirectly to Engineering Team
No Sales Fluff10+
Protocols Shipped
$20M+
TVL Overall