Aptos Pushes Encrypted Mempool Upgrade to Protect Users From Frontrunning and Censorship

Aptos has introduced a proposal for a native Encrypted Mempool system that would allow users to submit transactions privately while still maintaining the speed and transparency of the network.

If approved through governance, Aptos said the feature would make it the first Layer 1 blockchain to offer built-in encrypted transaction submission directly at the protocol level.

Aptos Targets MEV Exploitation

The system is designed to protect users from frontrunning, censorship, and orderflow manipulation. Users would be able to send encrypted transactions with a single click, while all transaction data would still become visible on-chain after block confirmation.

Aptos said the proposal comes as decentralized exchange activity continues to grow rapidly. It added that DEX spot trading volumes regularly surpassed $200 billion per month in 2025 and averaged roughly $476 billion monthly during the third quarter. While decentralized exchanges removed reliance on centralized custody and settlement systems, Aptos noted that most blockchains still expose pending transactions before they are finalized, which allows validators and other network participants to observe and potentially exploit trading activity before execution.

According to Aptos, this visibility has contributed to the rise of the MEV market, where validators and traders profit by reordering or exploiting pending transactions. The proposed Encrypted Mempool aims to eliminate that exposure by ensuring transaction intent remains confidential until execution while preserving the network’s same security assumptions.

Aptos Labs explained that the system relies on threshold cryptography and a distributed key generation process that occurs before each validator epoch. Transactions are submitted as encrypted payloads, and validators collectively decrypt them only after a block has been ordered. The company added that traditional encrypted transaction systems face major scalability issues because validators must individually communicate and process partial decryptions for every encrypted transaction. This ends up creating heavy communication, computation, and latency costs across the network.

To solve this problem, its research team developed a batched threshold decryption scheme that allows validators to generate a single partial decryption for an entire batch of encrypted transactions instead of handling them individually. Aptos said this significantly reduces communication and computation overhead while allowing most processing work to happen in advance.

The company further revealed that the system prevents replay attacks, removes the need for users to compete for encryption slots, and avoids transaction resubmissions. Aptos said the encrypted mempool integrates directly into the network’s consensus protocol and introduces minimal additional latency.