Claim◆Article

This could not be a better time for AI to get good at cryptanalysis, unless AI succeeds in undermining the hard problems themselves or we live in Impagliazzo's Minicrypt.

Green argues that the transition period is an ideal window for AI-driven cryptanalysis, with the caveat that AI might instead undermine the foundational hard problems or we inhabit a Minicrypt world where they don't exist. ✦ AI generated

Matthew Green · Simon Willison's Weblog · 2026-07-29 · original ↗

Share card ↓Export .md Export .txt

If there was ever a perfect time for a massive new public cryptanalysis capability to come on line, we're in it. So unless AIs succeed in undermining all of our hard problems altogether (or we live in Impagliazzo's Minicrypt) then this could not be a better time for AI to get good at cryptanalysis.

Read full article ↗excerpt · fair-use quotation

✳ AI ✳ cryptanalysis ✳ Impagliazzo's Minicrypt

Around this claim

Mechanism · 1

We are in the midst of a historic transition from traditional public-key algorithms based on EC and RSA to post-quantum algorithms based on novel problems.Matthew Green · Simon Willison's Weblog · conf 75%

In practice · 1

Autonomous exploit development by frontier AI agents is no longer a hypothetical capability.ExploitGym authors (UC Berkeley, Max Planck Institute, UC Santa Barbara, Arizona State) · Simon Willison's Weblog · conf 70%

Extends · 1

In the best case, AI cryptanalysis gives us real confidence in the problems we have identified and makes the cryptanalysis literature considerably more robust.Matthew Green · Simon Willison's Weblog · conf 70%