rotative
techcrunch ai · June 13, 2026

Anthropic touted its model as a bomb — now the government made it pull the plug

the u.s. government ordered anthropic to shut down its fable 5 and mythos 5 models worldwide, citing national security risks and a reported jailbreak.

read in 简体中文 · 繁體中文 · 日本語 · 한국어

anthropic was forced to shut down access to its fable 5 and mythos 5 models worldwide following a u.s. government directive issued on friday. the order, framed as an export control measure, was reportedly triggered by a narrow jailbreak of fable 5, which anthropic argues is a standard cybersecurity capability already present in models like openai's gpt-5.5. the shutdown is ironic for anthropic, which had restricted mythos due to its advanced security vulnerability finding abilities and positioned itself as the safety leader. the company criticized the move, warning that applying this standard broadly would effectively halt all new model deployments.

read at source ↗

← rotative