Claude Mythos Preview

Technology

GLM-5.3 is said to be able to build end-to-end cyber exploits autonomously and to have been released without robust misuse safeguards

Techmeme relays Anthropic's claim that GLM-5.3, like Claude Mythos Preview, can autonomously build end-to-end cyber exploits.

TE

1 Source, 9h ago, first seen 9h ago

AI

GLM-5.3 reportedly nears Mythos Preview on a controlled browser-exploit test

Anthropic reports 50 successful ExploitBench attempts out of 410 for GLM-5.3, versus 56 for Mythos Preview. Its separate safeguard figures measure simulated engagement, not successful attacks.

BB

EM

AK

26 Sources, 3h ago, updated 3h ago

AI

Anthropic Claude Mythos Preview Finds Cryptographic Weaknesses

Anthropic researchers employed the model to uncover vulnerabilities in secure digital signature schemes and symmetric ciphers.

SZ

LS

KA

26 Sources, 63d ago, first seen 63d ago

Claude Mythos Preview

Recent stories

GLM-5.3 is said to be able to build end-to-end cyber exploits autonomously and to have been released without robust misuse safeguards

GLM-5.3 reportedly nears Mythos Preview on a controlled browser-exploit test

Anthropic Claude Mythos Preview Finds Cryptographic Weaknesses

Claude Mythos Preview

Recent stories

GLM-5.3 is said to be able to build end-to-end cyber exploits autonomously and to have been released without robust misuse safeguards

GLM-5.3 reportedly nears Mythos Preview on a controlled browser-exploit test

Anthropic Claude Mythos Preview Finds Cryptographic Weaknesses