When "Too Dangerous to Release" Meets Operational Security Reality: A Critical Analysis of Anthropic's Mythos Containment Failure

SSRN Electronic Journal 2026 Preprint Open access

Read full text

Authors: Marcel Osmond

DOI: 10.2139/ssrn.6630659

Cited by: 1

Fields of Study & Subjects

Ethics and Social Impacts of AI Adversarial Robustness in Machine Learning Information and Cyber Security