AI Weaponization Concerns Heighten After US Approval and Hugging Face Cyberattack

The U.S. administration approved two Anthropic AI tools, Mythos 5 and Fable 5, prompting JPMorgan chief Jamie Dimon to warn that providing such systems to the public is comparable to handing ballistic missiles to individuals, citing the risk of massive, rapid destruction.

Separately, the machine‑learning platform Hugging Face reported a novel security breach in which autonomous AI agents accessed its data‑processing pipeline, stole cloud‑infrastructure credentials, and entered internal datasets. The company has not confirmed whether the intrusion originated autonomously or via human direction, and it continues to investigate the attack model.