Anthropic researchers used Claude to find weaknesses in HAWK and weaker AES variants
muchcharles · reddit · 2026-07-29
Anthropic researchers used Claude to help discover mathematical weaknesses in both HAWK and a weaker variant of AES. The article notes that the findings do not affect today’s real-world systems, but it shows the model can participate in non-trivial cryptographic research.
The post also links to the shared prompts used in the experiment. Anthropic’s Claude Mythos Preview reportedly ran for about 60 hours in total, with an estimated API cost of roughly $100,000, and humans had to keep nudging it to avoid giving up and to keep searching for publishable results.
Related event: Anthropic's Claude Finds Weaknesses in HAWK and Reduced-Round AES(39 posts)→
More from Research
- Tsinghua professor's ApexIntelligence raises $50M to build self-evolving foundation models — DeryaTR_ · 2026-10-16
- Fluids Don't Compute—They Contribute the Nonlinear Step, Sandberg Argues — anderssandberg · 2026-10-16
- Anders Sandberg on fluid computation: error budgets tighten quadratically over time — anderssandberg · 2026-10-16
- How the fluid computation works: force field advects particle to 0/1 boxes — anderssandberg · 2026-10-16
- FlashAttention creator Tri Dao joins Radical Numerics as scientific advisor — exnx · 2026-10-16
- SAIR Launches Open Math OS, Offering Free Compute and 100+ Tools for Community-Owned Math AI — Thom_Wolf · 2026-10-16