I predict that we will continue to see more like this about post-quantum schemes: "Despite HAWK having survived two rounds of expert human review over a period of two years, Mythos was able to improve the best-known attack on it in just 60 hours of work—effectively cutting its key strength in half." (see https://www.anthropic.com/research/discovering-cryptographic-weaknesses ), but (and i think this is the common wisdom) very little such things with the hash-based PQC schemes. This is common wisdom because non-AI research was finding holes in other post quantum schemes a lot, for years. The problem with hash-based schemes is they're kind of shitty - precisely because they eschew any structure, we can't do anything nice with them.

(Disclaimer: my post 100% sure lacks important nuance, I have not studied this field...)