Jonathan Passerat-Palmbach, a senior research scientist at Flashbots and visiting scholar at Imperial College, discussed the academic evolution and practical limitations of encrypted mempools at the EthCC[9] conference. According to Foresight News, he highlighted the progression from early committee-based threshold encryption schemes to batched threshold encryption and the Beast-Mev scheme, which combines silent settings with batch decryption. Despite these developments, no current solution meets all the protocol-level requirements such as silent settings, non-interactive decryption, small keys, and small ciphertexts, preventing the cryptographic aspects of encrypted mempools from being directly integrated into the Ethereum protocol.
The economic implications pose a more significant challenge. The blind ordering mechanism used in encrypted mempools, where transactions are ordered while encrypted, leads to three main issues: searchers submitting numerous blind arbitrage bundles, resulting in on-chain spam; reduced MEV subsidies making geographically remote validators less competitive, thus increasing centralization; and a decline in user execution quality, with significant transaction price deviations and higher rollback probabilities. Jonathan noted that the timing design of decrypting after execution hinders the effective operation of auction mechanisms, which are optimal for handling MEV incentives. He suggested programmable privacy as an alternative, allowing a balance between full encryption and full transparency, currently achievable through TEE and potentially through more advanced solutions like FHE in the future. He concluded that privacy should not exist independently of economic incentives and should continue to be experimented with outside the protocol rather than being hastily integrated.
Weatherly