SentinelLABS releases VerifiedIDA to make AI-driven reverse engineering easier to check and reuse

SentinelLABS tested LLMs reverse engineering binaries and annotating IDA Pro databases without safeguards. The models produced inaccurate labels, left findings unrecorded, lost track of unfinished work and stopped analysis early. The team proposed a standard requiring each annotation to be traceable to evidence, outstanding work to be tracked, and independent review, and open-sourced VerifiedIDA as a reference implementation for IDA Pro.

For IT people · Written by AI from 1 source · Wrong?