STORY · FORSKNING_
LLMs will not break symmetric encryption
Anthropic announced in July that their LLM Claude Mythos discovered new cryptanalytic attacks, including a key recovery attack on HAWK (a post-quantum signature scheme) and an improved attack on 7-round AES-128. At the same time, Anthropic developed CryptanalysisBench, a benchmark for LLM-assisted cryptanalysis. An analysis concludes that LLMs will not be able to break established symmetric encryption algorithms like AES, ChaCha or SHA-3.
WHY IT MATTERS
The findings show that LLMs can play a role in cryptographic research by finding weaker or alternative attacks, but simultaneously demonstrate that established algorithms are robust enough to resist LLM-assisted cryptanalysis due to their targeted design without mathematical structure.
SOURCES
MACHINE-GENERATED SUMMARY This summary is written by machine from the sources below. We sort and explain — but we are a way into the field, not the final word. Check the source when something matters to you.