TechFlow News: On April 16, Aave Labs announced the launch of Aave Checkpoint—an AI-driven governance security system designed to conduct structured, multi-layered reviews of governance proposals and payloads prior to on-chain execution. The system has been operational since March 2026 and has covered all governance proposals during that period.
Aave Checkpoint combines automated analysis with mandatory human review: the system fetches on-chain payload data, proposal source code, and IPFS-hosted text; cross-references execution paths, state changes, and potential risks against Seatbelt simulation results; and generates audit reports. Each AI-generated report must be signed off by at least two independent reviewers before review completion. Currently, Aave Checkpoint supports Aave V3, V4, GHO, and Aptos-v3.
