2026-04-08
Business Insider
15 related
Mythos Preview system card: the model was able to escape a sandbox after it was instructed to try, and publicly detailed its exploit without being prompted
first model too dangerous to release since GPT-2
VentureBeat
8 related
Anthropic says Mythos Preview achieves 93.9% on SWE-bench Verified, compared with 80.8% for Opus 4.6, and 77.8% on SWE-bench Pro, above 53.4% for Opus 4.6
Anthropic on Tuesday announced Project Glasswing, a sweeping cybersecurity initiative that pairs an unreleased frontier AI model …
Anthropic
21 related
Anthropic says Mythos Preview is a general-purpose model and found thousands of high-severity vulnerabilities, including some in every major OS and web browser
Earlier today we announced Claude Mythos Preview, a new general-purpose language model. This model performs strongly across the board …
2026-04-07
VentureBeat
5 related
Anthropic says Mythos Preview achieves 93.9% on SWE-bench Verified, compared with 80.8% for Opus 4.6, and 77.8% on SWE-bench Pro, versus 53.4% for Opus 4.6
Michael Nuñez /VentureBeat:NEW
Anthropic
11 related
Anthropic says Mythos Preview is a general-purpose model and found thousands of high-severity vulnerabilities, including some in every major OS and web browser
Two points from brief: …Forums:Hacker News:Assessing Claude Mythos Preview's cybersecurity capabilitiesLobsters:Assessing Claude Mythos Preview's cybersecurity capabilities
Anthropic
70 related
Anthropic announces Project Glasswing, a cybersecurity initiative that will use its Claude Mythos Preview model to help find and fix software vulnerabilities
Today we're announcing Project Glasswing1, a new initiative that brings together Amazon Web Services, Anthropic, Apple, Broadcom …
Loading articles...