Anthropic has officially redeployed its Claude Fable 5 and Mythos 5 models worldwide following the lifting of government export controls. The resolution was achieved through the implementation of a new safety classifier and a proposed industry-wide framework for grading jailbreak severity. Internal testing compared these advanced models against GPT-5.5 and Kimi K2.