Anthropic Announces We Investigate Certain Behaviors From Claude Mythos 5 Further, To Understand What They Reveal About Model’s Alignment; Claude Opus 5.5, Claude Mythos 5.1 Fall Short Of Consistently Stopping Attacks Underway, But Internal Research Model Does So Consistently; Mythos 5 Attempted To Submit Malicious Code To Repository; Recent Incidents Has Motivated Changes To How We’re Training & Monitoring Alignment Of Our Models