Aptos Labs Targets Frontrunning with Encrypted Mempool Proposal

The proposed system functions by allowing users to submit encrypted transactions, which validators then sequence without sight of the underlying order details. According to a research paper authored by Rex Fernando, Guru-Vamsi Policharla, Andrei Tonkikh, and Zhuolun Xiang, this design integrates directly into the network's Byzantine fault-tolerant consensus protocol. Testing indicates this added layer of confidentiality introduces approximately 27 milliseconds of latency, representing a 14% increase over the current baseline performance.

To manage the computational load, validators share decryption keys during each epoch, utilizing batched processing to ensure efficiency. This initiative aligns with Aptos Foundation's broader $50 million investment strategy aimed at bolstering trading infrastructure and AI-driven applications. Among the primary beneficiaries is the on-chain order book Decibel, which has already processed over $1 billion in volume. As the network expands its institutional footprint—supported by tokenization integrations from firms like DigiShares—the encrypted mempool represents a critical step toward providing the privacy guarantees required by professional traders.