Account Abstraction Research Report: ETH Account System Evolution and the Landscape for the Next Five Years

Summary

Ethereum underwent a major upgrade called Fusaka on December 3, 2025. This is the third milestone update since The Merge and Dencun upgrades, aiming to significantly enhance network scalability, reduce transaction costs, and optimize node operation efficiency. It also focuses on upgrading and optimizing account abstraction (AA) related functionalities. As a major upgrade to Ethereum’s account system, AA seeks to address the fundamental security and user experience bottlenecks of the “private key = account” model in the EOA era, enabling on-chain accounts to have programmable, recoverable, and permission-controlled modern capabilities. The implementation of ERC-4337 has accelerated the formation of the smart account ecosystem, but challenges such as high costs, unclear business models, ecosystem fragmentation, and cross-chain limitations have caused adoption to lag behind technological advances. With rollup cost reductions, BLS signature aggregation, and EIP-7702 providing seamless upgrade paths, AA is gradually entering a scalable application phase. Over the next five years, AA will become the “high-end smart layer” of on-chain accounts, coexisting with EOAs and the x402 interoperability protocol, jointly driving Web3 from a niche to mass adoption, and laying the foundational hub for a unified internet account system.

1. The Development History and Capabilities of AA Accounts

Ethereum’s major upgrade Fusaka on December 3, 2025, marks the third milestone since The Merge and Dencun, aiming to significantly improve scalability, lower transaction costs, and optimize node efficiency, with a focus on upgrading account abstraction (AA). In Ethereum’s system, the evolution of account structures is central to on-chain user experience, asset security, and industry upgrades. The familiar dual-account system of EOA (Externally-Owned Account) and CA (Contract Account), introduced in 2015, is a legacy of early design. However, as user numbers surpass tens of millions and Web3 increasingly handles asset custody and user operations between 2023–2025, this system reveals growing structural bottlenecks. These bottlenecks limit industry expansion, user growth, and real-world application deployment. AA emerged to address these inherent structural flaws, endowing on-chain accounts with “modern financial-grade” security, experience, and autonomy, ultimately becoming a trusted asset infrastructure for global users. The core reason for these bottlenecks is that EOAs embed the “private key = asset” security model at the protocol level—simple in engineering but a major obstacle to large-scale adoption in practice.

EOA operations resemble a “mechanized assembly line,” rather than the “one-click execution” familiar to modern internet users. Additionally, EOAs cannot implement fine-grained permission controls: no daily limits, multi-signature rules, parent-child accounts, partial permission freezes, or automation strategies. EOAs are like master keys containing all assets and permissions; once leaked, assets and permissions are fully exposed.

Consequently, the Ethereum community began rethinking “what should an account be,” and AA’s philosophy directly addresses this: an account should be “code,” not “private key.” Under the AA paradigm, accounts can be programmed, verified, recovered, and upgraded. In other words, restrictions hardcoded into the EOA architecture can be abstracted away. Wallets are no longer just signature containers but can become “smart accounts” with logic, strategies, and permission systems. The concept of account abstraction did not emerge overnight but involved a long design debate, including proposals like EIP-86, EIP-2938, ERC-4337, and the latest EIP-7702. EIP-86 and 2938 require modifications to the Ethereum consensus layer, making them difficult to implement; ERC-4337’s brilliance lies in building AA as a “parallel system,” using UserOperation and Bundler to bypass protocol modifications, enabling seamless integration into the existing Ethereum ecosystem. Essentially, ERC-4337 creates a new parallel channel alongside the transaction mempool, allowing users to submit UserOperations instead of transactions. Bundlers then bundle, simulate, and aggregate these operations before submitting them to the EntryPoint contract, enabling features like proactive contract account transactions, batch execution, atomic operations, and multi-signature verification. Although this increases engineering complexity, it provides a practical path to fully enable AA without hard forks. Between 2024–2025, Vitalik proposed EIP-7702 to further facilitate natural conversion between EOAs and AA, though ecosystem support remains necessary. AA’s significance extends beyond fixing EOA’s structural flaws; it introduces a “generational leap” in experience, security, and cost efficiency. First, security: AA enables programmable permission systems—social recovery, multi-sig, parent-child accounts, whitelists, spending limits, temporary keys, and permission freezes—making wallets more flexible and secure. The “single point of failure” mode of EOAs is thoroughly mitigated, greatly enhancing security. Cost-wise, after introducing Paymaster, users can pay gas with any ERC-20 token, or have projects pay on their behalf, creating a “fee-less” experience. AA also supports batch execution and transaction aggregation, reducing signature counts and failure costs, thus lowering overall complexity and expense. On the user experience front, AA makes Web3 interactions closer to Web2: users can perform combined operations with one click, without understanding nonce, gas, or signature order; new users can create wallets without mnemonic phrases, using biometrics, local recovery, or email verification; complex on-chain logic like strategy trading, automated liquidation, or scheduled execution can be embedded into accounts, enabling “smart product” capabilities.

The ultimate vision of AA is to transform blockchain from a “techie experiment system” into a “universal account infrastructure” for global users. If the past decade’s Web3 bottleneck was “keys as accounts,” the next decade’s breakthrough will be a “programmatic account” paradigm. AA is not just an upgrade of wallets but a rewrite of on-chain interaction logic; it enhances user experience, lowers development barriers, and allows DApps to design workflows, define permissions, and build trustless security systems like Web2 products. As ERC-4337 ecosystem matures in 2024–2025, with the rise of Bundlers, Paymasters, AA wallets, and modular security plugins, account abstraction is shifting from “concept” to “infrastructure.” Just as mobile evolution from Web1.0 to Web2.0 spawned super apps and trillion-dollar industries, the implementation of AA could become a foundational driver for Web3’s next exponential growth. The limitations of the EOA era are gradually being dismantled, and AA is leading the industry toward a safer, more flexible, and more user-friendly on-chain world.

2. The Future and Challenges of AA Accounts

Account abstraction (AA) re-emerged as a core narrative in Ethereum’s ecosystem during 2023–2025, but after initial enthusiasm, its structural challenges have become apparent. The long-term outlook remains optimistic—AA promises a generational leap in security, usability, and automation, replacing the “private key = account” model of EOAs. However, in practice, the implementation of ERC-4337 has faced repeated skepticism, often dismissed as “big talk but little action.” From industry structure, cost models, ecosystem collaboration, and protocol competition perspectives, the prospects and difficulties of AA are intertwined, representing both the future of blockchain account systems and the complexity of protocol upgrade paths.

On the cost side, gas is the primary obstacle. Compared to 21,000 gas for a standard EOA transaction, UserOperation on mainnet averages around 42,000 gas—nearly double. This is not waste but structural: validation calls like validateUserOp, state access in EntryPoint, reading wallet contract bytecode, logging, initCode deployment, and data encoding all add overhead. Each step involves additional on-chain computation. Theoretically, running complex logic within contract wallets is correct, as accounts should be programmable, verifiable, and controllable; but Ethereum’s expensive L1 resources mean all these designs translate into costs, which severely hinder adoption. Many potential users and projects are deterred. On the business model front, Paymaster faces unclear ROI issues. Paymasters, which pay gas for users in exchange for growth or value retention, lack mechanisms to accurately measure “gas paid → new users → retention and conversion.” Most wallets or DApps initially rely on subsidies to attract users, but once subsidies end, user migration is easy, making network effects hard to establish. The Web3 ecosystem also lacks the “advertising, retention, and traffic loop” of Web2, making Paymaster’s investments often unrewarded and unsustainable. Therefore, slow promotion of AA is not purely a technical issue but a lack of commercial incentives. The market does not pay for ideas but for profits. Ecosystem fragmentation worsens the 4337 dilemma: the full stack includes EntryPoint, Bundler, Paymaster, Wallet Contract, and Aggregator, with different implementations across wallets and chains. Due to the complexity, UserOperations are processed via Bundler simulation and aggregation, leading to incompatibilities among implementations. Wallet incompatibility, high DApp integration costs, and complex on-chain testing force projects to reassess the cost-benefit ratio. EOAs are simple but primitive; AA is advanced but initially causes ecosystem fragmentation. For most small and medium DApps, supporting 4337 yields no immediate benefit and adds technical overhead, leading to the “prefer not to adopt” outcome.

Cross-chain capability is another systemic weakness. ERC-4337 is fundamentally an EVM layer account system upgrade relying on EntryPoint, UserOp, and EVM validation logic, making it inherently difficult to extend to non-EVM chains. Achieving a unified multi-chain experience would require additional layers, multiple EntryPoints, repeated validation, and cross-chain messaging, exponentially increasing costs and complexity. Web3 is already fragmented across chains, and AA cannot unify accounts across chains, undermining the vision of “Web3’s unified account standard.” On different chains, user accounts cannot be frictionlessly mapped, greatly reducing AA’s scalable value. Despite these structural challenges, AA remains a promising direction because the evolution of next-generation blockchain infrastructure aligns naturally with AA’s principles—especially with the rise of Layer 2 (Rollup) solutions, which can structurally reduce AA’s costs. Data compression in ZK Rollups and Optimistic Rollups can cut gas costs by 70–90%, and batch UserOperations further lower per-operation on-chain costs. Thus, “Rollup + AA” is likely to be the dominant combination over the next 35 years, alleviating the cost burden on Ethereum mainnet. Concurrently, ERC-4337 continues to evolve, notably with the introduction of BLS signature aggregation. By aggregating multiple UserOperations into a single signature for batch execution, on-chain data volume drops significantly, increasing TPS and reducing gas consumption. This enhances on-chain throughput, transforming AA from a “wallet upgrade” into a “more efficient on-chain operation protocol.” Coupled with Rollup compression, core bottlenecks in performance and costs are being unlocked, making industry see its commercial viability. Additionally, Vitalik’s EIP-7702 offers a “temporary conversion” path from EOA to smart account, allowing users to enable AA instantly without migrating assets or changing wallets. EIP-7702 greatly reduces ecosystem congestion, enabling wallet providers to upgrade gradually without fundamental architecture changes, making AA nearly imperceptible to users. This marks a key turning point: AA no longer needs to “replace EOAs” but can coexist and gradually evolve alongside EOAs through progressive upgrades.

However, the biggest future challenge for AA comes from the rise of the x402 protocol in 2024–2025. Unlike AA, x402 functions more like an “Internet-level unified payment protocol,” using HTTP 402 as an entry point to unify Web2 and Web3 payment interfaces. AA aims to abstract accounts within chains; x402 aims to abstract internet payments. AA targets Web3 users; x402’s potential audience is the entire internet. More importantly, x402 has a natural commercial closed loop: Providers and Facilitators can directly charge for payment services, creating clear market incentives. ERC-8004 becomes a “tool protocol” within the x402 framework, not a foundational infrastructure requiring full network migration, making it easier to promote than AA. AA must persuade the ecosystem to migrate to its system, while x402 adopts existing internet behaviors, giving it a commercial advantage. Therefore, AA’s prospects are clear but challenging. The tension between elegant technology and industry realities is profound: AA’s envisioned future is better, but requires overcoming costs, incentives, ecosystem fragmentation, and protocol competition. As rollup ecosystems mature, signature aggregation advances, and EIP-7702 enables compatibility, AA’s costs and interoperability issues will ease, but business models and cross-chain capabilities still need breakthroughs. The key in the coming years is whether the industry can find a natural way for AA to diffuse. The future of AA belongs to ecosystems that can connect “protocol capabilities → product experience → business value,” not just technical implementers. It may not be the easiest to promote, but it remains the most promising solution to reshape on-chain account systems.

3. Investment Value and Future Outlook of AA Accounts

Account abstraction (AA) is shifting from a “revolutionary technology concept” to a “structural infrastructure upgrade,” with its investment value evolving from early narrative hype to a comprehensive assessment of engineering implementation, ecosystem collaboration, and commercial sustainability. Over the next five years, AA will not become the entire Web3’s unified entry point nor replace EOAs as the standard account system, but it will firmly exist at the high-end of wallets and account systems, serving as the core of “smart accounts,” deeply embedded in on-chain interaction experiences and transaction execution in the Rollup era. For investors, AA’s value is not short-term user explosion but a “long-term infrastructure investment opportunity” akin to classic internet infrastructure.

From a structural trend perspective, AA’s position will significantly rise with the adoption of EIP-7702.7702 allows EOAs to temporarily become smart accounts within a single transaction, meaning existing wallets do not need forced migration or asset restructuring. Users can enjoy AA’s permission controls, social recovery, multi-sig, and automation without changing wallets, mnemonic phrases, or migrating assets. This “painless upgrade” makes AA’s adoption curve smoother, incentivizing wallet providers to incorporate it into core architecture. Over the next three to five years, we are more likely to see coexistence and integration of EOAs and AA rather than outright replacement.

The main deployment of AA will be within Rollup ecosystems. As zkSync, Scroll, StarkNet, and Base become mainstream execution environments, the cost issues of AA will be naturally absorbed by Rollup data compression, reducing gas costs by 70–90% compared to L1. BLS signature aggregation and batch UserOperations will further decrease on-chain data size, transforming AA account operations from “costly but advanced” to “advanced and affordable.” This means the investment value lies not in L1 AA but in deep integration with Rollup-based AA wallets, Paymasters, and Bundler infrastructure. This direction offers tangible engineering value—reducing actual on-chain costs and driving real adoption. Industry-wise, AA’s investment opportunities mainly focus on four infrastructure sectors: smart contract wallets, Paymaster providers, Bundler infrastructure, and AA-compatible L2s. Wallets like Safe, Argent, OKX Web3 Wallet, imToken (AA version), and Zerodev are the most promising “ecosystem targets,” enabling a transition from “key wallets” to “smart account wallets” with modular architecture, social recovery, multi-sig, and automation, ensuring strong user retention. Paymasters are among the most commercially promising components, acting as bridges between gas subsidies and user growth. Although current models are immature, in richer Rollup environments, they could become “on-chain growth engines”: projects subsidizing gas for high-value users, running campaigns, and whitelist strategies, creating marketing effects similar to Web2 advertising. Projects like Stackup and Pimlico are worth watching. Bundlers, as the execution layer of AA, are foundational infrastructure akin to “transaction packaging logistics.” Companies like Biconomy and Alchemy’s AA infrastructure will benefit from ERC-4337 ecosystem growth. While Bundlers do not directly interface with users, their scalable revenue models could make them a “low-volatility, large-scale” on-chain infrastructure investment.

Meanwhile, the future five years will see AA face competition and complementarity from the x402 protocol. Unlike AA, x402 is more like an “Internet-level unified payment protocol,” using HTTP 402 as an entry point to unify Web2 and Web3 payment interfaces, with cross-chain capabilities and a clear commercial closed loop (Provider + Facilitator fee models). ERC-8004 becomes a plugin within the x402 framework rather than a foundational protocol, making it easier to promote. From an investment perspective, AA’s value lies in on-chain account intelligence, while x402’s value is in connecting the entire internet’s payment experience. Both will coexist and complement each other, not compete directly.

In summary, over the next five years, AA will form the “middle layer infrastructure” of Ethereum and Rollup ecosystems: the bottom layer remains EOAs (weakened but present), the middle layer is smart accounts (AA), and the top layer is the unified interoperable network of x402. AA’s user base will grow steadily with on-chain transaction volume, automation needs, and asset custody demands. In a world increasingly migrating on-chain, AA is a highly deterministic structural investment; in a world of declining Rollup costs, it is a “realized future”; and in an internet coexisting with x402, it is a core force shaping the on-chain account system.

4. Conclusion

The core value of AA lies in transforming Ethereum’s account system from the original “private key = account” model to a modern “account = program” paradigm. It fills a critical gap in the migration from Web2 to Web3, making secure, recoverable, and programmable wallets possible. Despite structural bottlenecks like high costs, weak business loops, and cross-chain limitations, AA has become a foundational infrastructure for on-chain experience upgrades. In the future, AA will exist as a high-end account layer long-term, not as the sole standard; x402 will complement cross-chain and payment connectivity. Together, they will push Web3 from a niche techie domain toward mass adoption and lay the key groundwork for a “unified internet account” system.