Kofi Ndaikate has emerged as a cornerstone in the rapidly shifting landscape of decentralized finance, where the battle between market integrity and predatory automation is fought every second. With a deep background in fintech regulation and blockchain architecture, Ndaikate has spent years dissecting the mechanics of Maximal Extractable Value (MEV) and its impact on global liquidity. As Aptos rolls out its Encrypted Mempool on mainnet, his perspective offers a rare bridge between high-level protocol design and the gritty reality of the trading floor. His insights are particularly timely as the industry seeks to move beyond the “dark forest” of transparent mempools toward a more secure, institutional-grade environment.
In this conversation, we explore the intricate mechanics of how transaction encryption effectively blinds frontrunning bots without compromising the speed of the network. We dive into the staggering economic data surrounding sandwich attacks on other major chains and analyze the cost-benefit trade-offs for traders who must now weigh higher gas fees against the security of their execution. Our discussion also touches upon the broader ecosystem strategy, including the $50 million commitment toward building robust financial infrastructure and the multi-layered approach to on-chain confidentiality that Aptos is currently spearheading.
How does the technical architecture of encrypting transaction instructions before they reach the network fundamentally disrupt the strategies employed by frontrunning bots?
The fundamental power of a frontrunning bot lies in its ability to see what you are doing before you actually do it, effectively operating like a gambler who can see everyone else’s cards at the poker table. By launching the Encrypted Mempool, we are essentially turning those cards face down until the very moment the pot is awarded. When a user initiates a trade through a compatible wallet like Petra, the specific instructions—such as which assets are being swapped or the exact smart contract functions being called—are encrypted directly on the user’s device. This means that as the transaction travels to the network, the validators and any external observers can see the sender’s address and the network fee, but the core “intent” remains a black box. It is only after the validators have reached a consensus on the transaction order that they collectively reveal the contents for execution. Because this decryption requires a stake-weighted majority of the validator set, no single rogue actor can peek inside to jump the line, which effectively neuters the advantage that these bots rely on to extract profit from unsuspecting traders.
Given the massive scale of automated manipulation seen on other blockchains, how does the recent data on sandwich attacks highlight the urgency for this kind of protocol-level defense?
The numbers we have seen recently are truly staggering and serve as a wake-up call for anyone involved in high-frequency on-chain trading. When you look at the investigation by Bitquery, which identified roughly 6.9 million sandwich attacks on Solana over an 11-month period ending this October, you realize this isn’t just a minor nuisance; it’s a systemic drain on capital. Those attacks targeted approximately 2.2 million individual wallets and managed to extract a jaw-dropping $29.7 million before accounting for their own trading costs. What is even more alarming is the sheer velocity of this growth, with daily attacks skyrocketing from 3,400 in late 2025 to over 51,900 in September of this year alone. While a huge chunk of that—about 96%—was concentrated in the volatile memecoin markets like those found on Pump.fun, the predatory tactics remain the same. Without an encrypted mempool, every large order is essentially a target, and this new feature is the shield that prevents that $29.7 million from being siphoned out of the pockets of legitimate users and into the hands of a few thousand bot operators.
For a trader or a developer, what are the practical trade-offs when choosing to opt-in to this encryption, particularly regarding the increased costs and network performance?
There is no such thing as a free lunch in cryptography, and the Encrypted Mempool does introduce specific economic and performance costs that users need to factor into their strategies. For starters, the network requires a minimum gas price of 200 octas per unit for these transactions, which is exactly twice the base minimum of the network. On top of that, you are looking at an additional 375 gas units just to handle the decryption of the transaction payload, which is separate from the gas used for the actual execution. In terms of speed, our testing with 50 validators showed that this process adds about 20 milliseconds of latency per block. While that might sound like a lot in the world of high-frequency trading, you have to remember that the standard AptosBFT consensus already finalizes blocks in under a second, so the delay is nearly imperceptible to a human user. For an institutional trader moving significant volume or a platform like Decibel, which has already cleared over $1 billion in cumulative volume, paying a bit more in gas and waiting an extra fraction of a second is a very small price to pay to ensure an order isn’t picked apart by bots.
How does this native implementation by the existing validator set compare to other industry attempts to solve the MEV problem, such as private transaction builders?
The beauty of the Aptos implementation is that it doesn’t rely on “side-cars” or external committees that might introduce new trust assumptions or points of failure. If you look at how Ethereum handles this, many users turn to Flashbots Protect, which routes transactions through private infrastructure rather than the public mempool. Other chains, like Gnosis, have integrated Shutter technology which uses its own set of encryption mechanics. Aptos, however, has built this directly into the protocol’s DNA, utilizing the same validators that are already responsible for securing the entire blockchain. This means the security of your encrypted transaction is tied directly to the stake-weighted majority of the network itself, rather than a third-party builder or a separate committee of “key-holders.” It’s a much more cohesive approach that keeps the power within the existing consensus mechanism, and since it’s based on the AIP-144 proposal, it has been rigorously vetted, including an independent security review by ZKSecurity to ensure the math holds up under pressure.
What is your forecast for the adoption of these privacy-centric features as the ecosystem continues to expand its reach into institutional and automated trading?
I believe we are entering an era where on-chain confidentiality will no longer be seen as an optional “luxury” but as a foundational requirement for any serious financial activity. We are already seeing the pieces fall into place with the “three layers” of confidentiality now live: Keyless Accounts for identity, Confidential APT for hiding transfer amounts, and now the Encrypted Mempool for protecting intent. When you combine these tools with the more than $50 million that the Aptos Foundation and Labs have committed to developing trading infrastructure and AI agents, you can see a future where automated systems can operate with a level of privacy that was previously impossible on a public ledger. Even though the native token is currently facing some market pressure—dropping 11.5% to around $0.7995 recently—the underlying utility of these features is what will drive long-term value. As more applications like Decibel and Shelby integrate these tools, the “dark forest” of blockchain will gradually become a transparent and fair marketplace where the best strategy wins, not the fastest bot. We are building the rails for a multi-billion dollar economy that can finally promise its participants that their data and their trades are truly their own.
What is your forecast for the future of MEV-resistant trading environments?
The landscape of 2027 and beyond will likely view transparent, unencrypted mempools as a relic of an era where we simply didn’t know better, much like the early days of the internet before HTTPS became the standard. As institutional volume continues to migrate on-chain, I expect the demand for native, validator-led encryption to become the industry benchmark because it removes the “trust” factor currently required by private relays. We will see a significant shift where the most successful DeFi protocols are those that default to privacy, protecting their users from the $29.7 million in annual losses we’ve seen elsewhere. Ultimately, the successful integration of these technologies will force MEV bots to evolve from predatory “sandwiching” toward more constructive roles, such as providing efficient arbitrage that actually benefits price discovery rather than just draining user wallets. This transition will be the catalyst that finally allows decentralized exchanges to compete head-to-head with the execution quality of centralized financial giants.
