Aptos Launches Encrypted Mempool, Blocking Sandwich Attacks

Aptos Mainnet Launches Optional Encrypted Mempool for Enhanced Transaction Privacy

Aptos, a prominent blockchain platform, announced on Monday the official launch of its Encrypted Mempool on the mainnet. This feature positions Aptos among a select group of mainstream public blockchains to integrate transaction privacy directly into its consensus layer. Users can opt to enable this functionality, ensuring that transaction details remain encrypted until block ordering is finalized, thereby preventing bots from exploiting pending orders for sandwich attacks or front-running.

Currently, the mempools of most mainstream public blockchains are entirely transparent. Any node can view the specifics of upcoming transactions, including the functions called, assets and amounts involved, and the addresses of interacting contracts. MEV (Maximal Extractable Value) bots leverage this visibility to insert their own orders before or after user transactions, profiting by buying low and selling high at the expense of legitimate traders.

Data from Bitquery investigations reveals that sandwich attacks alone generated $29.7 million in revenue for bots on Solana over an 11-month period, involving nearly 6.9 million attacks and 2.2 million wallets. Losses on Ethereum are even more substantial: according to EigenPhi, traders lost approximately $60 million to sandwich attacks in the 12 months leading up to October 2025. This figure does not account for other MEV strategies such as front-running and backrunning.

The operational logic of Aptos’s Encrypted Mempool involves four key steps. Users, through compatible wallets or applications, encrypt their transaction details on their local devices before submission. After encryption, only the sender’s address and gas fee are visible; core information such as assets, amounts, and contract interactions are masked.

Encrypted transactions then enter the mempool for queuing. Validators sort these transactions but cannot see their content, a process referred to as “blind sorting.” Once the sorting order is determined, all validators collectively decrypt these transactions before execution and on-chain recording. Throughout this entire process, no single node can independently decrypt pending transactions, thus preventing malicious actions by insiders.

According to Aptos developer documentation, the minimum gas price for encrypted transactions is double the network’s baseline minimum. In test environments with 50 nodes, this feature introduced approximately 20 milliseconds of additional latency per block. This cost is exchanged for the privacy of transaction details, a premium considered acceptable for institutional large orders or high-frequency trading strategies.

Encrypted mempools are not an entirely new concept. Gnosis Chain implemented an encrypted mempool in July 2024 via the Shutter protocol, and Ethereum’s Flashbots Protect routes transactions privately to block builders, bypassing the public mempool. Both solutions have proven effective but rely on additional infrastructure: Shutter utilizes an independent decryption committee, while Flashbots depends on trusted builders.

Aptos’s design differs by having existing validators handle both sorting and decryption simultaneously. This eliminates the need for third-party nodes or additional trust assumptions, offering a more unified architecture. However, it also means that the security ceiling is directly determined by the scale and honesty of the validator community.

Data indicates that MEV attacks are no longer a fringe issue. The figures of $60 million on Ethereum and $29.7 million on Solana demonstrate that this is a mature and systematic industry, rather than isolated incidents. By integrating an encrypted mempool into the public chain layer, Aptos signifies that a new wave of Layer 1 projects are designing transaction privacy as fundamental infrastructure, rather than an afterthought for dApps or third-party tools.

Future observations will focus on two key areas. Firstly, adoption rates: will institutional investors, AI agents, and large-volume traders genuinely utilize this feature? If only a small number of users adopt it, the volume of encrypted transactions may be insufficient to validate its economic viability and security. Secondly, regulatory implications: encrypted transaction details could create a gray area for anti-money laundering (AML) and law enforcement investigations, particularly when decryption occurs after transaction ordering. The question remains whether regulators can effectively trace these transactions. These two points will determine whether encrypted mempools become a mainstream standard or remain a niche functionality.