Expert Explains AI Verification Crisis: Math is Checkable, Cryptanalysis is Not

matthew_d_green · x · 2026-07-29

Cryptography expert Matthew Green further explains why verifying AI-generated results is so difficult.

He notes that in many high-profile public math results, the outcome is typically a simple but hard-to-find counterexample that can be tested via basic calculation or machine-checked proofs. However, this verification approach fails for a lot of non-practical cryptanalysis, making AI-generated false proofs incredibly hard to spot.

Related event: Experts Highlight Challenges in Verifying AI-Generated Results(2 posts)→

Original post →

More from Models

Models channel →