Garand from the Stellar Development Foundation explains how a $25M MEV attack on Ethereum exploited proof-of-stake's fundamental security flaws, and contrasts this with Stellar's proof-of-agreement consensus model that requires validators to be trusted by existing network members.
Garand, a protocol engineer at the Stellar Development Foundation, analyzes a recent $25 million MEV attack on Ethereum where rogue validators exploited flashbots despite having their stakes slashed. He explains that proof-of-stake's security model fails because validator deposits are too small compared to the assets being traded, and anyone can become a validator by paying an entry fee. Stellar uses proof-of-agreement instead, where new validators must be invited and trusted by existing validators, preventing anonymous malicious actors from joining the network. This approach eliminates the class of attacks seen on Ethereum while maintaining decentralization without a security council.